Criminal justice research begins with a question. A researcher might ask whether a particular policing strategy reduces crime, whether pretrial detention affects case outcomes, or whether a rehabilitation program reduces recidivism. The quality of the answer depends entirely on how the question is investigated.

A research question identifies what the researcher wants to know. A hypothesis is a specific, testable prediction derived from theory or prior evidence. Not all criminal justice research is hypothesis-driven — descriptive research aims to document patterns without testing causal claims — but the distinction matters because it shapes how findings should be interpreted.

Study design determines what conclusions are warranted. In an experimental design, participants are randomly assigned to treatment and comparison conditions. Randomization is the most powerful tool available for establishing causation because it distributes both measured and unmeasured differences between groups. The Kansas City Preventive Patrol Experiment, the Minneapolis Domestic Violence Experiment, and the Philadelphia Foot Patrol Experiment are landmark examples of randomized experiments in policing.

Random assignment is frequently impossible in criminal justice. Researchers cannot randomly assign people to poverty, incarceration, or criminal records. Quasi-experimental designs attempt to approximate experimental conditions using matched comparison groups, interrupted time series, difference-in-differences analyses, or regression discontinuity designs. These approaches can produce credible causal estimates but require careful attention to the assumptions they rest on.

Observational research — surveys, interviews, ethnographies, and archival analysis — does not attempt to manipulate conditions. It describes what exists. Observational studies are essential for understanding phenomena that cannot be experimentally manipulated, but they are more vulnerable to confounding: the possibility that an observed association reflects a third variable rather than a causal relationship.

Sampling determines whose experiences a study captures. A probability sample gives every member of the target population a known chance of selection, allowing researchers to generalize findings to the broader population within calculable margins of error. Convenience samples — volunteers, available records, accessible populations — are common in criminal justice research but limit generalizability. Students should always ask: who is in this sample, and who is not?

Measurement is the process of assigning numbers or categories to observations. Validity refers to whether a measure actually captures what it claims to capture. Reliability refers to whether a measure produces consistent results. A measure can be reliable without being valid — a scale that consistently reads five pounds too heavy is reliable but not valid. Crime statistics, for example, measure reported crime, not all crime. The gap between the two is the dark figure of crime.

Data collection methods each carry distinct advantages and limitations. Surveys can reach large samples efficiently but depend on respondents' willingness and ability to answer accurately. Interviews allow depth and follow-up but are time-intensive and subject to interviewer effects. Official records — arrest data, court filings, prison admissions — are comprehensive but reflect organizational decisions as much as underlying behavior. Observation provides direct access to behavior but is resource-intensive and raises questions about reactivity.

Research ethics in criminal justice is governed by the Belmont Report's three principles: respect for persons, beneficence, and justice. Respect for persons requires informed consent — participants must understand what they are agreeing to and must be free to decline or withdraw. Beneficence requires that researchers minimize harm and maximize benefit. Justice requires that the burdens and benefits of research be distributed fairly, not concentrated among vulnerable populations who bear the costs while others receive the benefits.

Institutional Review Boards review research proposals involving human subjects to ensure these principles are applied. Research involving prisoners, juveniles, or other vulnerable populations requires heightened scrutiny because the capacity for voluntary consent may be compromised by institutional power dynamics. The history of research abuses — including the Tuskegee syphilis study and coercive interrogation research — explains why these protections exist.

National Research Council. (2004). Fairness and Effectiveness in Policing: The Evidence. National Academies Press. https://doi.org/10.17226/10419

Sherman, L. W., & Weisburd, D. (1995). General deterrent effects of police patrol in crime 'hot spots': A randomized, controlled trial. Justice Quarterly, 12(4), 625–648.

Weisburd, D., & Majmundar, M. K. (Eds.). (2018). Proactive Policing: Effects on Crime and Communities. National Academies Press. https://doi.org/10.17226/24928

National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. (1979). The Belmont Report. U.S. Department of Health and Human Services. https://www.hhs.gov/ohrp/regulations-and-policy/belmont-report/index.html

Maxfield, M. G., & Babbie, E. R. (2018). Research Methods for Criminal Justice and Criminology (8th ed.). Cengage Learning.

Lum, C., & Nagin, D. S. (2017). Reinventing American policing. Crime and Justice, 46(1), 339–393.