ABSTRACT
The previous chapters presented mathematical equations to calculate sample sizes to achieve a desired power level in a study with hierarchical data. Sample sizes were derived to calculate the required number of clusters given a fixed cluster size or calculate the required cluster size given a fixed number of clusters. Attention was also paid to designs where neither sample size was fixed beforehand and a cost function was used to find the least expensive design to achieve a desired power level, or to derive the design with the highest power level on a fixed budget.
