Аннотация:Sample size requirements are given for intervention trials in which the unit of randomization is the cluster. Cases dealt with include the comparison of two means and the comparison of two proportions. Required modifications to the standard statistical analyses are also discussed.
