In experimental settings where data are collected with the goal of making inferences about some aspects of an underlying population it is always important to design the study in such a way as to obtain as much useful information as possible while minimizing the overall cost of the experiment. This is particularly true when the initial step in collecting these data is to select the particular units from the finite or infinite population on which measurements are to be taken. In this context, the goal of minimizing experimental cost is most often equivalent to minimizing the sample size while still achieving the desired accuracy of the inferences that follow.