Cluster sampling

Explanations > Social Research > Sampling > Cluster sampling

Use | Method | Example | Discussion | See also

Use

Use when the studied population is spread across a wide area such that simple random sampling would be difficult to implement in accessing the selected sample.

Method

Divide the population up into a set of different coherent areas.

Randomly select areas to assess.

Access all subjects in the selected areas. If you cannot do this, select a significant random sample and use the same selection rules in each cluster.

Example

In a study of the opinions of homeless across a country, rather than study a few homeless people in all towns, a number of towns are selected and a significant number of homeless people are interviewed in each one.

Discussion

Sometimes the biggest problem with sampling is being able to reach your targets, and having them are spread out over a large geographic area is a common experience.

Even when you have selected a cluster, you are unlikely to be able to access everyone in that cluster (you are unlikely, for example, to be able to interview everyone in a selected town). The practical answer is to select a significant and similar sample in each cluster. For example if you are going to interview people in clothes shops, you should do this at the same time on the same weekday in each cluster (you would, after all, likely get different results interviewing 9am Monday morning from if you did it on Saturday afternoon).

Cluster sampling may be combined with other forms of sampling, for example proportionate quota sampling, to ensure sub-groups are fully represented.

A risk with cluster sampling is that some geographic areas can have different characteristics, for example affluence or political bias.

Cluster sampling is also called area sampling.