Nonequivalent Group Designs

The nonequivalent group design is a quasiexperiment that draws comparisons among groups of participants who are assigned to different treatments in a nonrandom fashion. Selection differences are likely to bias the estimates of the treatment effects. Analysis strategies and design features for taking account of this bias are described. These analysis strategies and design features impose different assumptions about the nature of selection. Unfortunately, researchers will seldom have enough information to know which set of assumptions is correct in any given situation. In general, the recommended approach is to minimize the bias as much as possible by making the nonequivalent groups as similar as possible.