Common Study Designs for Criminal Justice

Criminal Justice

There are a large number of study designs that can be used to better understand issues in social sciences, including criminal justice (USC, 2014). Before a study design can be applied to a particular phenomenon or problem, however, they must be understood in terms of the strengths and weaknesses. In addition, most study designs will not be appropriate for a given situation and research goal. To better understand the strengths and weaknesses of particular study designs and appropriate applications this essay will discuss five that would be useful in criminal justice research. Two of these methods will then be used to explore how they could be used to study a professional practice issue in criminal justice.

Study Designs

Exploratory Study Design

Exploratory research is used when little is understood about a particular phenomenon and there is the possibility of future research using more quantitative methods (USC, 2014). The goals are primarily to provide a comprehensive description of a problem, suggest possible theories that would be applicable, and evaluate whether the problem would be amenable to additional study in the future. In other words, the exploratory study design would be useful for understanding whether a new, untested policing strategy may be having an impact on a particular crime statistic like domestic violence. For example, if an urban policing agency recently created a community policing unit it might be useful to interview the officers assigned to the unit, along with members of the public who the officers have encountered, to better understand whether the community policing strategy is producing better outcomes. The findings from an exploratory study cannot be generalized to the wider population because of the small sample size and conclusions about causality cannot be made.

Case Study Design

One of the study designs that could be used following exploratory research is the case study, because it allows researchers to narrow the research focus, thereby providing an opportunity to understand a phenomenon in greater depth (USC, 2014). Since only a limited number of subjects are included this study design is useful for investigating and characterizing issues that occur only rarely. Because of the small sample sizes, however, the findings cannot be generalized to the wider population and conclusions about cause and effect relationships cannot be made. In addition, there is no procedure that can control for selection and researcher biases. An example of a case study would be a detailed analysis of an encounter between community police officers and members of the public who were involved in a domestic dispute.

Cohort Study Design

The next logical step in researching the efficacy of community policing on domestic violence outcomes would be a cohort study. This study design uses a larger sample size than exploratory and case study designs and the subject included in the study all share some trait (USC, 2014). For example, all domestic violence statistics generated for a community policing neighborhood could be collected for a year and compared to same statistics collected for a neighborhood without community policing. Cohort studies therefore provide quantitative data and can make tentative conclusions about cause and effect (USC, 2014). The limitations are significant, however, and include the inability to control for confounding factors. For example, differences in the socioeconomic status of the two neighborhoods being compared could be having a significant impact on domestic violence statistics.

Experimental Study Design

To address the often significant limitations of a cohort study design an experimental study design is typically used (USC, 2014). The experimental study design utilizes procedures that attempt to control as many variables as possible and may involve randomization of study subjects to two or more conditions. If the ethical issues are set aside for the purpose of discussion, then an example of an experimental study design would be randomizing police responses to domestic violence calls. A comparison could be made on outcomes depending on whether a regular patrol or community policing unit responded to the call. A number of demographic variables would have to be collected for each encounter so that potential confounding factors could be controlled for. For example, the socioeconomic status of the entire neighborhood and the residents involved in a domestic violence response could be controlled for, thereby increasing the validity of the findings. In addition, experimental studies often provide findings about cause and effect relationships that can be generalized to the wider population.

Meta-Analysis Study Design

If enough cohort and experimental studies have been completed investigating the same phenomenon of interest then the findings can be analyzed and combined in a meta-analysis to provide insights otherwise unavailable (USC, 2014). A meta-analysis is only possible, however, if the studies included used similar research methods and provided enough information about the data to render it amenable to statistical analysis. For example, if five policing agencies studied the impact of community policing on domestic violence outcomes using a cohort study design then conducting a meta-analysis could be informative. The conclusions reached in a meta-analysis would be weakened if the findings from the included studies differed in significant ways, but strengthened if there was general agreement among most or all studies. The overall effect of a meta-analysis is to increase the sample size and therefore the statistical power of the analysis.

Policing in a Post-Stop-n-Frisk Era