A SPLIT MATRIX
Comparing Studies of Moderate and Strong Methods
A. Quasi-experiments of B. Studies using stronger quasi and
moderate quality also randomized experimentation
This separation shows that there are many fewer studies available of high quality that tell us about the effectiveness of police interventions. However, as Weisburd et al. (2001) have shown, higher quality criminal justice evaluations are less likely to show positive effects, which means that we are more likely to see positive effects in Figure A. A few other things stand out in this division - higher quality studies that show positive effects are most consistently found in the proactive, micro place-based region. Figure B also allows tells us, with high certainty, that individual strategies are much less promising, and in some cases, harmful. Finally, notice that neighborhood and group-based studies almost completely disappear when only looking at the highest quality studies. If we wish to continue using such strategies, then better information must be generated at these units of analysis.