THE BASIC CONCEPT

We will start by describing randomisation tests for assessing differences between treatment groups in a standard comparative experiment. It may help understanding to consider first the assumptions made in the more common t-test or analysis of variance.

Let us suppose that there are two groups. The t-test assumes that the two groups arose by drawing samples from two Normally distributed populations, and that we are investigating whether these populations differ in their mean. We ask how likely we are to have obtained the observed difference between the samples if there is no difference between the population means.

The randomisation test, on the other hand, assumes that some initial set of individuals were randomly allocated to two treatment groups. We ask how likely it is that a randomisation would produce a difference as big as that obtained.

Randomisation tests. 02.00 Page : 2 of 14.