Qualitative research is often divided into either experimental or quasi-experimental research. Experimental research typically involves a control group and an experimental group performing an action and a variation of an action, respectively, in order to gauge the differences in results that should occur. A quasi-experimental design is an experiment along the same lines, but there are some unique differences that set it apart. In this paper, the differences between experimental and quasi-experimental research will be examined (Cozby, 2008).
Experimental research is strong in internal validity - it allows for X to lead to Y, because evidence is provided for this proposition, setting the program apart from whatever other causes can lead to the outcome. This makes it much more internally valid. The control group and the experimental group form a two-group experiment, which forms the basis for experimental research. Quasi-experimental research, on the other hand, is much less reliable in terms of internal validity, as the assignment of subjects is not random. Instead, the experimenter must make sure to pick intact groups that approximate control and treatment groups, like two classrooms that are similar enough in type to be comparable (Cozby, 2008).
The nonequivalent groups design is the most common type of quasi-experimental design, and its makes statistical analysis of the experiment much more complicated. Both types of design require a pretest and posttest, and have two-group designs. Fortunately, the lack of randomization allows for more versatility and range in these quasi-experimental designs. The regression discontinuity design (RD) variety of quasi-experimental research involves the establishment of a cutoff that sets assignment of subjects ot one group or another based on certain criteria. This allows for a bit more internal validity by providing comparison units in greater quantities, providing a sophistication of measure on par with experimental design.
Probabilistic equivalence is the primary aspect that sets experimental research apart from other research designs. This is the idea that a pair of groups will perform along the same lines, provided they are measured for long enough. Random assignment is the key to successful experimental research - by randomizing the results, bias is minimized and it will make sure that probabilistic equivalence is ensured. As long as the sample is large enough to be statistically significant, an experimental design can bring about clear, positive results (Cozby, 2008).
The main problem with experimental design is that a true experimental design can be hard to carry out in a real world situation; the high internal validity often comes from the artificial situation that is set up to achieve maximum control and limit bias. Quasi-experimental design is much easier to accomplish, but the internal results are not quite as valid. In the end, it is recommended to perform an experimental design in an artificial environment due to the high internal validity, but often quasi-experimental research is often the only option for a real-world experiment.
References
Cozby, P. C. (2008). Methods in behavioral research. Mountian View: Mayfield Pub. Co.
Creswell, J. W. (2009). Research design: qualitative, quantitative, and mixed methods approaches (3rd ed.). Thousand Oaks: SAGE Publications.
Trochim, W. M. (2001). Research methods knowledge base (2nd ed.). Cincinatti: Atomic Dog Pub.
