This study aimed to address the lack of evidence for the practical implementation of procedural justice goals by conducting an experimental evaluation of a training program aimed at promoting the use of procedural justice by officers in the Seattle Police Department (SPD). The primary analytic focus was on tractable, policy relevant, measures of police performance and safety in the field.
During the field experiment, researchers identified officers involved in the top 12th percentile of predicted risk in their precinct based on the High Risk Circumstances (HRC) model. Officers were then randomly assigned to either a control or a treatment group. 241 officers were assigned to treatment and participated in engagements. In the experimental engagement group, officers were trained in procedural justice techniques.
At the conclusion of the engagement, officers were asked to fill out a confidential comment card about the experience, which was be mailed directly to the Police Foundation. Researchers then used administrative data from the Seattle Police Department to compare the post-engagement behavior of officers who work in behavioral hot spots on a number of dimensions during one and six week periods (short run and long run).
Officers in the treatment and control groups were compared based on the number and type of incidents that they responded to, their probability of using force or having a complaint filed against them, and the frequency with which they resolved incidents with a formal arrest, rather than a less disciplinary measure.
The High Risk Circumstances (HRC) model developed by the researchers was used to identify incidents in the top 12th percentile of predicted risk in their precinct on a bi-weekly basis. The officers involved in these incidents were identified and the top percentile of risk were randomly assigned to either a control or a treatment group. The experiment had 12 waves over the course of six months, and officers typically participated in meetings within 11 days of being notified of their selection. A total of 241 officers were assigned to treatment and a total of 221 procedural justice meetings were held.
Cross-sectional ad-hoc follow-up
All officers in the Seattle, Washington Police Department in 2013.
administrative records data
This study includes one excel file and one stata file.
- Comment Card database2.xlsx: This file includes 67 cases of police officer responses on comment cards, providing feedback on their supervisory meetings. The 11 variables include questions about whether the meeting helped the police officer to remember the event in question, whether the police officer felt respected or criticized, and whether the supervisor meeting provided a useful context for the officer to provide feedback to the supervisor.
- FinalRegressSet2.dta: This file includes administrative data used to assess study outcomes for 31,300 cases. The 61 variables include patrol district, data collection dates, predicted risk score, information about officer events (including arrests, reports, assists, citations, and warnings) both pre- and post- treatment, and the calculated differences between the same.
A total of 241 officers were assigned to treatment, and while union and IRB rules mandated that officer participation in the engagements was voluntary, there was 91.7 percent compliance rate.
Likert-type scales were used.