Figure 1. Simulation of a small cluster-randomised trial, first four runs.

Four plots each showing two groups. Each group consists of two sets of 10 points, close together. Each plot has a P value, P<0.0001, P=0.4, P=0.06, P<0.0001 respectively. The first plot shows group 1 with observations tending to be greater than group 2, the fourth plot shows group 2 greater than group 1.

Back to "Cluster randomised trials in the medical literature"

This page is maintained by Martin Bland

Last updated: 6 November, 2003