ABSTRACT
True experimental designs involve random assignment to treatment conditions and are excellent for exploring cause-and-effect relationships. However, it is often not possible to assign participants at random. In this case, Quasi-experimental designs, which are of intermediate value for exploring cause and effect, should be used. Quasi-experimental design can take many forms. This topic covers the two most common designs. Design 4 is a widely used, quasi-experimental design that has two intact groups. This is the nonequivalent control group design. This design is subject to the threats of mortality, selection, and interaction of selection with other threats such as history. Thus, it is far from ideal. Even though participants are not assigned to groups at random in Design 4, researchers often use matching to increase the internal validity of the results. The participants in Design 5 are, in effect, serving as both the control participants and the experimental participants.
