The results of previous systematic reviews of Neighborhood Watch are divided according to the conclusions drawn. Titus (1984) concluded that Neighborhood Watch was effective, but noted that the research methods were weak. Husain (1990) concluded that there was little evidence that Neighborhood Watch worked. Based on the four studies meeting their selection criteria, Sherman and Eck (2002) concluded that Neighborhood Watch was ineffective in reducing crime. The strongest finding of this review relates to the mean effect size estimate produced by the meta-analysis. This indicated that, across all eligible studies combined, Neighborhood Watch was associated with a reduction in crime. It is not immediately clear why Neighborhood Watch is associated with a reduction in crime; however, it is possible that the reductions were associated with some of the essential features of the Neighborhood Watch programs as discussed earlier. Neighborhood Watch might serve to increase surveillance, reduce opportunities, and enhance informal social control. Unfortunately, this kind of information is not provided in the majority of evaluations and the precise reasons for the reduction cannot be determined.