Experimental Criminology

The Experimental Design

Virtually every undergraduate textbook on research methods describes experimental designs as the “gold standard” of research methodology. While there are many variations on the experimental design, the classic experimental design involves an independent and a dependent variable, random assignment to experimental and control groups, and pre- and posttesting. In the classic experiment, the independent variable can be thought of as the cause, and the dependent variable as the effect. The researcher believes that the independent variable causes changes in the dependent variable and conducts an experiment to test this belief. To conduct a classic experiment, the researcher selects a target population and randomly assigns subjects to either an experimental or a control group. Random assignment of subjects to the treatment and control [Page 657]groups ensures equivalency of the groups. The researcher then conducts a pretest by measuring both groups on the dependent variable. Following the pretest, the experimental group receives the experimental stimulus while the control group does not. The researcher then remeasures both groups on the dependent variable. Changes in the dependent variable seen in the experimental group (but not in the control group) are presumed to result from exposure to the experimental stimulus.

A criminal justice example might be helpful to demonstrate these points and clarify the definitions. Imagine a researcher interested in the effectiveness of juvenile “boot camps.” Specifically, the researcher decides to test the hypothesis that juvenile boot camps increase a juvenile delinquent's respect for authority. The researcher identifies a target population as juvenile delinquents and draws subjects for the experiment from that population. The researcher then randomly assigns juvenile delinquents to the experimental and control groups and pretests each group's respect for authority prior to any intervention. The researcher then exposes only the subjects in the experimental group to the experimental stimulus, in this case boot camp. Following boot camp, both the experimental and control groups are remeasured on the dependent variable (respect for authority). Changes in the dependent variable observed in the experimental group but not the control group are presumed to result from exposure to the experimental stimulus (the boot camp). If the experimental group exhibits a greater respect for authority in the posttest, the researcher has proved his or her hypothesis and demonstrated the positive effect of boot camps on respect for authority.

The advantages of the experimental designs are widely acknowledged. Random assignment in an experimental design reduces conscious (or subconscious) bias in the assignment of subjects and also controls for the possibility of chance affecting the outcome. Control groups, a key element of an experimental design, provide a distinct advantage. The control group allows the researcher to attribute any changes in the dependent variable observed only in the experimental group to the experimental stimulus. Without a control group, it would be almost impossible to know whether the experimental stimulus, or something else, affected the experimental group. Due to the unique features of experimental designs, these research designs more readily allow for “causal inferences” (a statement of the cause-and-effect relationship between the experimental stimulus and outcome). While researchers frequently acknowledge the benefits of experimental designs, they also recognize that there are often significant barriers to conducting such experiments. Some of these barriers are particularly salient in the field of criminal justice.

...