Crime Prevention and What Works
There is an enormous body of criminal justice literature on crime prevention efforts. However, little of this literature examines the impact of crime prevention strategies. Instead, much of the research describes different types of programs and the manner in which they are implemented. The research that does exist often is of such poor quality that it does not permit one to draw conclusions regarding the effectiveness of the program studied.
The scientific standards for inferring causation have been clearly established and can be used to evaluate the strength of evidence included in each program evaluation. The Maryland Report used a scale of 1 to 5 to summarize the scientific rigor of the studies examined. The scores generally reflect the level of confidence that can be placed in an evaluation's conclusions about cause and effect, with a score of 5 indicating the strongest evidence and a score of 1 considered so low in scientific rigor that the results were excluded from conclusions about a topic. Studies were evaluated by determining their scientific merit and the outcomes. The scientific method scores reflect the strength of the evidence about the effect of the programs on recidivism. The outcomes (direction and size of the effect) were evaluated based on differences between the treatment group, which received the intervention, and the control or comparison group, which did not receive the intervention.
A large body of research on corrections programming for juveniles is in agreement with Altschuler and Armstrong. However, the quality of much of this research is disappointingly poor. Many of the studies only describe the program being evaluated and give recidivism rates for the participants without providing any information on the rates for a comparable group of juveniles who did not participate. Therefore, it is impossible to draw conclusions about the impact of the program. Other research attempts to make comparisons between different groups of participants and nonparticipants. However, the research is so poorly designed (a score of 1 or 2 on the Maryland scale) that it is impossible to rule out alternative explanations for the outcome results.