In recent years, school districts across the country have implemented numerous programs aimed at reducing fear of crime, often without the benefit of theoretical insight and best practices awareness. The consequence was a lack of universal success at fear reduction. However, the Charlotte School Safety Program (CSSP) was designed to create an environment wherein students join with teachers and police officers in assuming shared responsibility for reducing delinquency and disorder. In addition, the CSSP emphasized the development of social competency skill and appropriate behavioral norms. An evaluation conducted in 1998 reported significantly lower levels of fear following program participation. This article examined findings from this evaluation of CSSP, which endorsed program continuation despite minimal attention to context. The article begins with a review of the original study; issues specific to the context questions were identified and employed in a multivariate model. This reexamination with minimal attention to context revealed different results. Preexisting differences were identified among groups for key variables, indicating possible selection bias that could lead to inaccurate interpretations of the problem-solving program’s impact on outcome variables such as students’ perceptions of safety. In addition, the analysis enabled a more conservative test of treatment effects and the identification of variables that could render the initial observed treatment effects false. Appendix and references