Policing Schools Strategies: A Systematic Search for Experimental and Quasi-Experimental Studies
UK National Police Improvement Agency and the George Mason University Center for Evidence-based Crime Policy
WestEd researchers undertook a systematic search of the literature to identify experimental and quasi-experimental evaluations of policing schools strategies.
Police have long implemented strategies at schools, and this study takes stock of the evidence that assesses the effectiveness of those strategies. Although police-taught prevention curricula like Drug Abuse Resistance Education (D.A.R.E.) have been the subject of many high-quality evaluations and at least two systematic reviews, the evidence concerning other police-led school strategies is less well-known. This study is an attempt to take stock of what is known.
This study resulted in a published report.
This study documents a systematic search to answer the following question:
- What experimental and quasi-experimental evaluations have been reported that assess the effectiveness of non-educational policing strategies and programs in schools?
Studies were included if they reported on a specific school-based strategy that heavily involved police and did not exclusively involve the police teaching a curriculum or program such as Drug Abuse Resistance Education (D.A.R.E.).
The review focused on studies of policing strategies conducted in or directly around primary/secondary schools (K-12th grade in the U.S. education system). Only those impact studies that used experimental or quasi-experimental design, had at least one outcome measure of school crime or disorder, and were available through December 2009 were eligible.
To locate published and unpublished reports, WestEd researchers conducted electronic searches of over 50 bibliographic databases. To supplement searches of electronic data bases, researchers checked bibliographies in over 100 prior systematic and narrative reviews, identified citations referenced in every retrieved evaluation report, and corresponded with over 25 researchers who have conducted policing research.
The searches identified a total of 11 quasi-experimental studies. Nine of the studies used a non-equivalent comparison group design without statistical controls or matching procedures. In short, 10 of the 11 studies would likely have received a "3" on the Maryland Scientific Methods Rating Scale, a common approach to classifying studies on the basis of internal validity.
If evidence rating criteria from the U.S. Department of Education's What Works Clearinghouse were applied, only one study would likely receive a grade of "Level 2" evidence (acceptable with reservations) and the other 10 studies would likely not meet WWC evidence screening criteria.
Based on the results of the searches described here, and the internal validity issues of the studies identified (acknowledged in almost all cases by the original investigators themselves), WestEd researchers would conclude that a full systematic review to guide police and school decision-making by informing them of the scientific evidence on "what works" is premature. Instead, such procedures should be delayed until randomized experiments and additional quasi-experiments are implemented to study non-educational police programs that target schools.
Researchers recommend randomized experiments, regression-discontinuity designs, and statistically equated comparison group designs for larger scale studies that involve a number of schools. Short, interrupted time series with comparison group designs would be appropriate for cases in which one treatment school with the policing program is being compared to a comparison school without it.