Aspects of statistics in studies of cell proliferation. I. Multistage sampling

In cell kinetic experiments, as in many other branches of biological science, we often sample at various stages of an experiment: we may for example take a sample of animals, then from each study a sample of sites, and from each site take replicate observations. This sampling process can be optimized to give maximum precision to an estimated quantity, but care must be taken in analysing data so gathered because the analysis depends on the precise sampling strategy.