Biased sample

In epidemiology or medical research, a sample of a group that does not equally represent the members of the group.