Subject variables and natural treatment are quasi-independent variables because they cannot be directly manipulated.
The most common quasi-experiment is a nonequivalent-control-group design in which experimental and control groups have been predetermined or predesignated by either an existing subject characteristic or an already-occurred natural treatment and are not created by random assignment.
Matching individuals or groups on a variable that is highly related to the dependent variable is commonly used to equate participants before the study.However, matching on a pretest measure can present confounds related to regression to the mean for interpreting posttest results.
A quasi-experimental approach can incorporate all of the techniques and methods (with the exception of random assignment and random sampling) of research design including longitudinal methods, correlational techniques for data analyses, and statistical tests of interaction for studies with two or more independent variables.
A quasi-experimental approach has proven helpful in psychological studies of media multitasking, traumatic memories, and multiculturalism.
Age used as a subject variable is confounded with numerous other factors that pose serious threats to internal validity. In developmental studies, a cross-sequential design helps to control for confound related to age. A cross-sequential design increases the internal validity of developmental research.
