Skip directly to content

Randomized Experiments

Randomized experiments have a long history at the Police Foundation. Click here to read Dr. David Weisburd’s reflections on the Police Foundation’s substantial contribution to the field of experimental police research.


Promoting Officer Integrity Through Early Engagement and Procedural Justice (Ongoing)

This experimental evaluation of a training program is aimed at promoting the use of procedural justice by officers in the Seattle Police Department. The innovation of the proposed training program is twofold. First, we apply insights from criminology and statistics to develop a new kind of early intervention system, which we call a High-Risk Circumstance (HRC) model. This model identifies officers working in behavioral hot spots, small geographic areas where police officers are more likely to be involved in problematic citizen encounters. Our HRC model will be calibrated using geographically identified incident level data from 2009 collected by the Seattle Police Department.


Near Repeat Burglary Intervention (Ongoing)

This project aims to correct the deficiency of actionable crime prevention strategies for police agencies by using the knowledge surrounding near repeat burglary to develop a crime prevention strategy. Broadly speaking, our research seeks to determine if knowledge about near repeat patterns of burglary can actually be used for crime prevention purposes. Within this framework, we are attempting to determine if raising awareness about crime issues and crime prevention techniques with the residents near burglary locations can reduce further burglary in the area.


Automatic Vehicle Location (AVL) Technology in Policing (2012)

This study will examine the reliability and utility of the use of AVL technology to quantify police presence.  If the data are reliable, the study would next examine the impact of police presence on crime in specific geographic areas in Dallas, Texas and, more important, test whether AVL can be an effective tool for deploying officers. The results of our study would provide police agencies with the capacity for using AVL technology to assess and adapt deployment patterns.  This approach would advance information-led policing nationwide by providing a technology-based strategy for crime prevention and reduction.


The Shift Length Experiment (2011)

Karen L. Amendola, David Weisburd, Edwin E. Hamilton, Greg Jones, & Meghan Slipka (December 2011)

Most law enforcement agencies have traditionally deployed their patrol officers based on a 40-hour workweek in which personnel work five consecutive, 8-hour shifts, followed by two days off. In recent years, however, an increasing number of agencies have moved to some variant of a compressed workweek (CWW) schedule in which officers work four 10-hour shifts per week or three 12-hour shifts (plus a time adjustment to make up the remaining four hours of the standard 40-hour workweek). While this trend towards CWWs has been moving apace, there have been few, if any, rigorous scientific studies examining the advantages and disadvantages associated with these work schedules for officers and their agencies.

This report presents the results of the first known comprehensive randomized experiment of CWWs in law enforcement. The Police Foundation experiment was designed to test the impacts of three shift lengths (8-, 10-, and 12-hour) on performance, health, safety, quality of life, sleep, fatigue, alertness, off-duty employment, and overtime among police. In addition to scientifically rigorous research design and methodology, the number of reliable outcome measures employed to analyze the impact of shift length, including departmental data, laboratory simulations and exercises, and previously validated self-report instruments, make this study one of the most comprehensive ever undertaken in this area. The experiment was conducted in the Detroit (MI) and Arlington (TX) Police Departments between January 2007 and June 2009.

The Police Foundation's shift length experiment has received the 2012 Outstanding Experimental Field Trial Award from the Division of Experimental Criminology/American Society of Criminology (ASC). The award recognizes a single research project or program that contributes significantly to criminological research and experimental science.


Risk-Focused Policing at Places: An Experimental Evaluation of the Communities that Care Program in Redlands, California (2008)

Drawing upon the risk and protective factors model and suggestions regarding the importance of places for understanding and targeting juvenile delinquency, the Redlands Police Department has developed a risk-focused policing approach to places (RFPP). Specifically, the Communities that Care Program (CTC) in Redlands used the risk and protective model developed by Hawkins and Catalano to identify census block groups where the risk of juvenile delinquency was high. The Redlands Police Department employed a problem-oriented policing approach to develop and implement strategies to reduce risk factors in these areas and reinforce protective factors. The program was aided by sophisticated crime mapping tools, which allowed the police department to draw from multiple data sources in defining the places for intervention, in analyzing problems, and in developing innovative solutions.

Throughout the intervention period, the Redlands Police Department developed and implemented community- and problem-oriented policing strategies in the thirteen block groups that were randomly assigned to the treatment condition. Innovative treatments directed at juveniles were not administered to the thirteen control block groups. Using a hierarchical linear modeling approach, our experimental data suggest that the risk-focused policing program in Redlands did not consistently reduce problematic behavior or impact upon student perceptions of risk and protective factors. We also do not find a statistically significant impact upon student perceptions of legitimacy of police actions, though juveniles overall in Redlands had strongly positive attitudes toward the police.

Our findings appear to contradict recent reviews of police intervention at problem places, which have shown overall strong crime prevention benefits. There are a number of factors that might have led to the disappointing outcomes of the evaluation, including the level of geographic focus of the project and the nature of the interventions that characterized the treatment. We conclude that risk-focused policing at places should be focused on more micro units of place and have greater fit between problems and treatments if it is to have strong programmatic outcomes.

See Risk‐Focused Policing at Places: An Experimental Evaluation, David Weisburd, Nancy A. Morris, Justin Ready, Justice Quarterly, Vol. 25, Iss. 1, 2008.


Preventing Repeat Incidents of Family Violence: A Randomized Field Test of a Second Responder Program in Redlands, California (2007)

Robert C. Davis, David Weisburd, and Edwin E. Hamilton (December 2007) Police Foundation Report

This field test, conducted with the cooperation of the Redlands, CA, Police Department, sought to vary one of the parameters thought to affect the impact of second response programs. Victims who called the Redlands police with a domestic abuse complaint were randomly assigned (a) to receive a second response within 24 hours, (b) to receive a second response within seven days, or (c) to receive no second response. A check of police records and surveys with victims six months after the initial complaint was called did not indicate any reduction in new abuse resulting from any second response condition. The current findings, coupled with earlier research results, strongly suggest that second response programs are at best ineffective in reducing the potential for new abuse and at worst may increase the likelihood of new abusive incidents. Implications for criminal justice policy are discussed.


Richmond's Second Responders: Partnering with Police Against Domestic Violence (2005)

Richmond's Second Responders: Partnering with Police Against Domestic Violence

This study's primary purpose was an evaluation of the Richmond, Virginia, Second Responders Program, a collaborative effort of the Richmond Department of Social Services and the Richmond Police Department. The program involved the introduction of social workers, based in the police precincts, to scenes of domestic violence to which police had responded, while the police were still on site. We also examined the researcher/practitioner partnership whose primary members were the Police Foundation, the Richmond Department of Social Services and the Richmond Police Department, with a view to contributing to our knowledge of this model.

The Second Responders experiment suggests that police perform better or are perceived to perform better in domestic violence situations when social service workers are present, and that women victims of domestic violence who receive an immediate social service response at the time of the incident may experience reduced incidence and prevalence of domestic violence.


The Metro-Dade Spouse Abuse Replication Project (1992)

Technical Report 

The Minneapolis Domestic Violence Experiment, had a dramatic impact on the awareness of police administrators of the problem of domestic violence and on the recognition that arrest might be an effective response to it. Recognizing the need for more generalizable results, the National Institute of Justice, which funded the original Minneapolis study, funded six additional domestic violence experiments. This report presents a summary of the Dade County experiments.

No effect attributable to follow-up by the special domestic violence unit was found in any type of analysis. On the other hand, significant effects were found, based on the initial victim interview, attributable to the arrest treatment with respect to both the prevalence and incidence of physical assaults against the original victim. Based on the results of the second-wave victim interview, significant effects were found attributable to arrest with respect to both prevalence and time to failure of attacks against the original victim; significance level of the effect on incidence was one decimal short of the standard .05 level. Although no significant arrest treatment effects were found with respect to subsequent offence reports, cases randomly assigned to the arrest conditions had significantly lower prevalence rates and times to failure than those assigned to the no arrest condition.


Catching Career Criminals: The Washington, DC, Repeat Offender Project (ROP), Police Foundation Report (1986)

Susan E. Martin and Lawrence W. Sherman (1986)

(Research Brief)

Two facts have been central to the debate on controlling street crime. First, a relatively small number of career criminals commit a disproportionate amount of crime. Second, prisons are overcrowded. The combination of these two factors has spurred interest in focusing police resources on catching the most active and dangerous chronic offenders. This is the report of the Police Foundation’s evaluation of the Washington, DC, Metropolitan Police Department’s project to apprehend career criminals.


Neighborhood Police Newsletters: Experiments in Newark and Houston (1985)

Executive Summary

Technical Report

Appendices

This report summarizes the results of evaluations of two experiments in which neighborhood police newsletters were mailed to households in two areas, one in Newark and one in Houston. The newsletters, published by the respective police departments, contained crime prevention advice, information about successful attempts tot prevent or solve crime and – in half the cases – local recorded crime statistics. The evaluation of these efforts consisted of controlled experiments conducted by the respective police departments and evaluated by the Police Foundation with funds provided by the National Institute of Justice.

Readership of the newsletters appeared to be relatively low. Only 53 to 63 percent of the persons interviewed recalled seeing the newsletter when shown an actual copy. Few measures of effect proved to be statistically significant. Such meager results pose a serious question about whether newsletters are an effective method of providing information to households whose members have less than high school educations.


The Houston Victim Recontact Experiment (1985)

Executive Summary

Technical Report

Appendices

This report summarizes an evaluation of a new program in which police officers re-contacted crime victims by telephone. These calls were intended to express police concern for their plight, to provide information about referral services, provide information about filing insurance claims, elicit additional information about the crime, and offer crime prevention literature. This program was intended to reduce victim’s fear of crime, increase their commitment to their community, and enhance their satisfaction with the quality of police service. The program was carried out in Houston, Texas. It was evaluated by the Police Foundation, with the support of the National Institute of Justice.

The experimental evaluation found no positive difference between victims who received the calls and those who did not. The evidence suggests the program may even have had a negative effect upon Hispanic and recently-immigrated Asian victims with limited facility with English.


Minneapolis Domestic Violence Experiment (1984)

Lawrence W. Sherman and Richard A. Berk (1984) Police Foundation Report

Arresting an assailant in a domestic violence case significantly reduces the likelihood of future violence. In the first scientifically controlled test of the effects of arrest for any crime, arrest was found to be the most effective of three standard responses used by police when responding to cases of domestic violence.


Arresting Shoplifters (1983)

(Research Brief)

In 1983, the Police Foundation undertook a study to determine whether arrest can indeed deter shoplifting. Researchers found that the shoplifters who were arrested were neither more nor less likely as a group to commit future shoplifting offenses than were those who had been detained by store security personnel and released.


The Newark Foot Patrol Experiment (1981)

(Research Brief)

The results of this experiment suggest that while foot patrol may not reduce crime, it reduces citizen fear of crime. Residents see their communities as safer and better places to live, and are more satisfied with police services.


The Cincinnati Team Policing Experiment: A Summary Report (1977)

Concludes that neighborhood team policing is hard to maintain but is a potentially useful alternative to traditional police patrol methods.


The Kansas City Preventive Patrol Experiment (1974)

George Kelling, Tony Pate, Duane Dieckman, and Charles Brown (1974)

This landmark experiment found that traditional routine patrol in marked police cars does not appear to affect the level of crime, nor does it affect the public’s feeling of security, thus refuting a tradition that had prevailed in policing for almost 150 years. The experiment demonstrated that urban police departments can successfully test patrol deployment strategies and that they can manipulate patrol resources without jeopardizing public safety.