Study
Bowers, Johnson, and Hirschfield (2004) used a quasi-experimental time series design to assess the impact of alley-gates on burglary rates in Liverpool, England. Treatment alley-gated areas were compared with the surrounding police department areas (PDAs), which were the comparison areas. The study used police data from the Merseyside County (England) Police Department. Data were obtained both for the preimplementation period (January 1998 to December 1999) and for the period of implementation (January 2000 through June 2003). This data were geocoded.
The study assessed for pregroup differences with a related samples t-test and found them to be stable over a historical preintervention period, thereby making postintervention period comparisons more valid.
The study used a related samples t-test to assess trends in the preintervention period, comparing the treatment areas with the PDAs. The test was not significant, indicating that the burglary concentrations in these areas did not differ prior to the intervention. This result suggested that the differences observed in the postevaluation period were attributable to the intervention and not some other variable(s).
This study ruled out the possibility of other crime prevention efforts occurring in Merseyside through the dose–response analysis. The dose–response analysis found significant relationships, which allows for more confident determination of causal effects of the alley gates. Specifically, regressions found the number of blocks secured was the strongest predictor of the reductions. Moreover, the R-squares were remarkably high (compared to most social science R-squares, which rarely reach 50 percent), with up to 55 percent of the variance in burglaries explained by the intervention. The study included a 12-month postimplementation evaluation period. A weighted displacement quotient (WDQ) was used to determine whether displacement (i.e., the relocation of crime) or spatial diffusion (i.e., a reduction in crime) occurred in seven buffer zones near the treatment areas.