Definition of Cluster Sampling

In cluster sampling, research makes clusters or small groups from a large geographically dispersed population. Each cluster is externally homogeneous but internally heterogeneous and each cluster is the mini-representation of the population.

After developing the clusters, the research uses a simple random sampling technique to select certain clusters while the remaining clusters are not represented in the research. There are two methods of sampling elements in cluster sampling: single-staged and double staged.

In the single staged method, all elements in each cluster are sampled while in the double staged method, the simple sampling method is first used to create subsamples of elements in each cluster.

Cluster sampling is different from stratified sampling. In stratified sampling, two groups are formed which are externally heterogeneous but internally homogeneous.