A design that lacks a true control group but involves comparing groups that are not randomly assigned.