ABSTRACT
The apparatus was reliable; the statistics correct; the conclusion wrong! Significant effects are meaningful only to the extent that the procedure was properly designed to eliminate other explanations. All too often, a lack of design at the onset forces investigators to resort to less powerful analyses and uncertain conclusions. Ergonomists generally use one of three design approaches: descriptive, experimental, and quasi-experimental. In addition, they encounter special problems doing applied research in the field. Research often begins by using descriptive techniques to identify the issues and parameters. These may then lead to specific questions to be answered with experimental designs that require more effort to gain their added power. Experimental results may then be tested in real-world situations that sometimes preclude true experimental designs. Though weaker, quasi-experiments can be strengthened by techniques that can also be applied to experimental and descriptive designs. Some definitions will help to explain these designs.
