Q: What is experimental design?

1 Answer to this question

First, let’s clarify what is and what is not experimental design.

Experimental research refers to an investigative approach which planfully uses manipulation and controlled testing to understand causal processes. Generally, one or more independent variables are manipulated to determine their effect upon a dependent variable.

Experimental research is considered complete when the researcher can confirm that a change in the dependent variable is solely due to the manipulation of the independent variable. The major benefit derived from experimental research designs is the ability to point to cause-and-effect (that is, “What was it that caused the change?”). A tightly controlled manipulation of variables allows for replication. Replication refers to confirmation of the experimental finding through repeating the experiment to corroborate the outcomes with other subjects.   

Non-experimental research refers to an investigation with a predictor variable that is not susceptible to manipulation by the experimenter. Typically, this means that other investigative devices must be used to interpret conclusions … such as a case study, correlational analyses, or a survey. Generally, there is far less confidence attributable to the outcomes of non-experimental research because of the uncertainty of what it was that caused any of the observable changes.

The most basic experimental design includes: (1) a pretest of the dependent variable (i.e., the feature that is being measured) in one or more members of the experimental group (the subjects); (2) the initiation of a treatment (the independent variable); and (3) a posttest to discern whether any resultant change can be attributed to the treatment intervention.

Experimental research design is the established 'gold-standard' in most fields of science and all high Impact Factor journals. This is because it establishes confidence that results are valid, have been interpreted without intentional or accidental bias, and since they are open to replication are more likely to be definitive.

Caven S. Mcloughlin, Ph.D., Editor-in-Chief Emeritus for a major commercial academic publisher.