In treatment comparisons, groups of people rather than individuals are sometimes allocated to receive the treatments being compared.
One reason for doing this is to avoid the inadvertent application of the treatment allocated to one of the treatment comparison groups to the other treatment comparison group (contamination).
Another reason is if the treatment is administered to a group rather than to individuals, like some public health treatments.
When clusters of people are allocated rather than individuals, care should be taken to avoid unit of analysis errors.
If you feel that this definition hasn't helped you to understand the term, click on our monkey to let us know.