Increasing the proportionate representation of a given group in a study in order to increase the statistical power of tests comparing associations among variables in that group with associations in other study groups. It is a procedure of stratified sampling involving the selection of a disproportionate number of participants from a particular group or stratum. Typically, this done with a population consisting of rare types. Drawing a random sample could result in just a few or none of such individuals. Oversampling is also used with a stratum that has a large variance.