ABSTRACT
Cluster sampling is another form of probability sampling. In this sampling method, researchers draw groups of participants instead of selecting individuals. Suppose a researcher wants to survey a sample of members of United Methodist churches throughout the United States. If the researcher obtains a membership list with addresses, he or she might draw a simple or stratified random sample of individuals and mail questionnaires to them. To help increase precision, the researcher could also stratify on geography and draw a random sample of clusters from each stratum, which would assure, for instance, that the North and South are represented in the correct proportions. Suppose, for instance, that church members in the South tend to be more conservative than members in the North. Members of any one cluster in the South are unlikely to reflect accurately the attitudes of all members nationally.
