Cross-Sectional Research Designs in Criminology and Criminal Justice
Cross-sectional research allows researchers to study a phenomenon or the relationship between variables at one point in time. In cross-sectional research data are collected once for each case (e.g., individual, neighborhood, city, state) at a single point in time. Cross-sectional research is most appropriate for studies that have descriptive or exploratory aims and in many cases are inadequate to assess causal processes due to an inability for researchers to establish temporal order—a necessary but not sufficient condition for causality. However, in some cases cross-sectional research may be adequate or even preferable to assess mediational models such as when the mediator variable is contemporaneous. Due to the nature of cross-sectional research, it is also not possible to utilize this approach to examine patterns or within-individual change in behavior over time. Alternatively, longitudinal designs typically follow the same cases over time and make observations (e.g., surveys, interviews) for each case at multiple time points. This approach allows for researchers to establish temporal order and assess both between and within-individual change over time. There is an ongoing debate among criminologists, particularly those interested in developmental processes and theories, regarding the worth of cross-sectional designs over longitudinal designs. This debate largely stems from the relationship between age and crime in what is typically referred to as the age-crime curve and the criminal career paradigm. That is, traditional cross-sectional research has found that crime peaks in mid- to late adolescence and then gradually declines across adult age groups. Despite the criticisms waged against cross-sectional research and the benefits of longitudinal research to establish temporal order, cross-sectional research designs still dominate criminology and criminal justice research. This may be due to some of the advantages of cross-sectional research and disadvantages of longitudinal research. For instance, cross-sectional research is relatively quick and affordable to carry out, making it a fairly accessible research design to conduct without funding and enables production of timely findings. Most importantly, researchers have developed unique methods of data collection and statistical analyses to overcome concerns of causality in cross-sectional research including retrospective or life history analyses, experimental survey designs, and repeated cross-sections or trend data. While these “fixes” do not allow researchers to assess within-individual change in outcomes (e.g., behaviors, attitudes, personality), they do address issues of temporal order and nonspuriousness within the cross-sectional design. The following article presents works that address various aspects of cross-sectional research and some examples of research utilizing this design.