Sampling for program evaluation

Generic calculators answer generic questions. These guides answer the ones evaluation teams actually face: how many schools a defensible learning survey needs, how sample size changes across evaluation designs, and how to monitor dozens of blocks without surveying all of them. Every page shows its formulas, and every number can be recomputed for your own program in the free Build mode.

Why a separate hub for evaluation?

Because program evaluation sampling is not survey sampling with different words. Evaluations compare arms, track panels through attrition, reach children through schools that inflate every sample, interview four stakeholder groups with four different logics, and must survive a skeptical reviewer at the end. Sampling Shala exists for exactly this: a free planning aid for education and social-sector teams, honest about the maths and about its own limits. The concept behind every formula lives in the Learn library.