ABSTRACT
Chapter 3 focuses on three different kinds of research designs used in the field of public administration: experimental, quasi-experimental, and nonexperimental. It familiarizes students with the concepts of internal and external validity, which help the researcher make logical conclusions and have confidence in them. It includes how to identify problems in a research design when attempting to make a causal inference (going beyond data description and establishing an explanation) from the data collected. The chapter addresses increased interest in behavioral public administration, which draws from the theories and methods of psychology and other behavioral science. It provides an example to illustrate how this new interest is closely related to the chapter’s topic.
The criteria for causality are explored in the chapter, as are the commonly used experimental and quasi-experimental designs employed to infer causality. In an experimental design, the researcher has control over the assignment of subjects to groups, the introduction of the independent variable, and the conditions under which research occurs. Quasi-experimental designs are missing one or more of these conditions. The chapter closes with a discussion of nonexperimental designs, those that do not have a control group and lack internal validity. Such designs are still of value to public administrators for information on whether an intervention or program has the intended results.
