ABSTRACT
This chapter discusses why program evaluators typically use samples in their studies rather than entire populations of people. It also explains the advantages and limitations of five commonly used evaluation sampling methods: random, systematic, stratified, cluster, and convenience sampling. The chapter also explains the difference between the unit of sampling (schools and hospitals) and the unit of data analysis (school or student; hospital or patients) and discusses how to calculate an appropriate sample size. A form for reporting on sampling strategy is also offered. This form is designed to show the logical relationships among the evaluation questions, standards, independent variables, sampling strata, inclusion and exclusion criteria, dependent variables, measures, sampling methods, and sample size.
