ABSTRACT
As we have seen in previous chapters, a key feature of a CRT is that observations on individuals in the same cluster are likely to be correlated. When data from a CRT are analysed, it is therefore essential to use statistical methods that take such correlations into account. If data on individuals are analysed using standard methods that assume independence of observations, the consequence will be that standard errors of estimates are underestimated and the signifi cance of any effects exaggerated.
