A sampling procedure wherein units of the sampling frame that are similar on one or more variables (typically geographic) are organized into larger groups (‘clusters’), and a sample of groups is selected. The selected groups contain the units to be included in the sample. The sample may include all units in the selected clusters or a sub-sample of units in each selected cluster. The ultimate purpose of this procedure is to reduce interviewer travel costs for in-person studies by producing distinct groups of elements where the elements within each group area are geographically close to one another.

« Back to Glossary Index