quasi-experimental research

Definition
Quasi-experimental design: A research design that incorporates a quasi-independent variable (an independent variable manipulation without random assignment).

Notes:

Quasi-experimental designs lack the critical feature of random assignment, but they are still useful for (cautiously) investigating cause and effect. (Suter 2012, pg. 3)

Weak preexperimental design can be greatly improved by the technique of matching, and as such would be called quasi-experimental designs, as explained later in this chapter (Suter 2012, pg. 13)

A quasi-experimental design incorporates an intervention, but it lacks the critical element of random assignment to groups. (Suter, 2012, pg. 21)

There still exists a treatment in quasiexperimental designs, in the sense that the researcher introduces a treatment or experimental program (sometimes called an intervention). (pg. 21)

This type (matched control group) of quasi-experimental design is used often, and its strength rests on how well the extraneous influences have been controlled through matching.

Quasi-experimental designs in research involve a manipulation (or an intervention) without the control aspect of random assignment to groups. They establish control in other ways, such as matching (e.g., matched comparison group designs), temporal patterns (e.g., time series designs), and counterbalancing (e.g., repeated measures designs). Matched comparison groups are especially prone to interpretation problems linked to nonequivalence (pg. 27)

Quasi-experimental designs incorporate treatments or interventions, but they lack the key element of random assignment to groups. This characteristic seriously compromises control in these designs. (pg. 41)