Abstract: | Making inferences about the spectrum of bacterial strains on a plate as a consequence of observing only one strain in a random sample of colonies from the plate is relatively straightforward whereas making inferences about the spectrum of strains within a human volunteer as a consequence of observing only one strain in a random sample from the plate requires some care. The issue of the appropriate number of colonies to sample from a plate to make inferences about strain uniqueness within a human volunteer is examined and illustrated with an example |