Research Designs
Research designs are the primary means of controlling for invalidity in research or addressing the casualty dilemma and are a method of eliminating competing causal factors. The traditional laboratory model acts as a model for other research designs. It is made up of three major components: equivalence, pretests and posttests, and experimental and control classes (Hagan, 2014, p.74).
Equivalence
Equivalence is the researcher's attempt to pick and assign participants to reference classes in such a manner that they are perceived to be identical in all significant respects (Hagan, 2014, p.74). Its aim is to eradicate selection bias. Randomization and mixing are used to achieve the equivalence of the classes to be evaluated. Randomization involves the random assignment of subjects from the same population to one of the other teams. When it is done, each item has an equal chance of being picked and an equal probability of being assigned to any of the comparison groups (Hagan, 2014, p.75). Matching, on the other hand, assures equivalence by selecting individuals for subsequent comparison groups based on certain matching characteristics such as age, sex, and race (Hagan, 2014, p.75). The groups are, therefore, similar concerning these traits. Combining both randomization and matching is possible. A good example is the Cambridge-Somerville study where 325 pairs of boys were matched depending on delinquency potential. One member of each pair was then randomly assigned treatment (Hagan, 2014, p.75). Most studies utilize matching instead of randomization to ensure equivalence. Purists, however, insist on randomization rather than matching (Hagan, 2014, p.76).
Pretests and Posttests
In the above case, the two groups can be assumed to be similar - since the common characteristic was delinquency. Both groups are then exposed to a pretest (observation) before their exposure to treatment and a posttest (measurement) after the treatment process. Any disparity between the two is believed to be brought about by the stimulus.
Experimental and Control Groups
At the end of the treatment process, the two categories form an experimental and a control group. The group exposed to treatment forms the experimental group, while the one unexposed to the stimulus or predictor valuable becomes the control group (Hagan, 2014, p.75). An excellent criminal justice example of the classic experimental design is the Candid Camera. Today, the use of for security purposes is routine. This was not the case in previous times when the importance of electronic devices was not comprehended. Their usefulness came to life in 1975 when the Seattle Police Department installed hidden cameras in stereo speaker boxes in seventy-five commercial establishments deemed to have high-risk potential robbery victims (Hagan, 2014, p.76). The businesses automatically became the experimental group because a similar group of companies did not receive the cameras.
The pretests in this case comprised collecting data on the percentage of robberies cleared by arrest and conviction rates before the project. In a compromising situation, the clerk would activate the camera by pulling the 'trip' bill from the cash drawer. A special project director would then print photographs of the robber.
In a posttest comparing the two groups, it was established that 55 percent of the robberies in experimental establishments were cleared by arrest while only 25 percent had the same outcome in the control firms (Hagan, 2014, p.76). Convicted robbers at the hidden camera sites stood at 48 percent compared with 19 percent at the control group institutions.
In conclusion, the classical experimental design has been utilized to bring about desired outcomes in several areas in the law enforcement arena. Equivalence, achieved via randomization and matching, provides a classical set of subjects who present an ideal sample for study purposes. As such, the classical experimental design remains superior to other designs. It is, therefore, poised to be utilized by numerous researchers in their various studies.
