RTC/OM HCBS Measurement Education Modules

Cluster sampling (large-scale studies)

  • Examples of clusters would be: geographic groups, provider agencies or other distinct information clusters, counties, regional offices
  • When establishing a cluster sample:
    • The population is first divided into clusters based on group membership.
    • Then, some clusters are randomly selected to recruit participants within.
    • Finally, a simple random sampling process is conducted within each of the randomly selected clusters.
  • Cluster sampling is often used when sampling all groups/clusters would not be feasible
  • Example: An HCBS provider with 94 group homes (clusters) serving adults with IDD selects 45 of the homes to randomly recruit participants within for a measurement effort.
Infographic illustrating what a cluster sample looks like. There are five groups or clusters of three people. Two of the groups have a magnifying glass over them to show they were the clusters selected for sampling within.