Character education programs support the development of positive character traits in children and adults. Effective violence prevention programs improve pro-social competencies and reduce negative behaviors in students by enhancing protective factors (strong bonds with teachers; clear rules of conduct that are consistently enforced) and targeting modifiable risk factors (inappropriate behavior; poor social coping skills). Comprehensive character education or prevention programs implemented in schools are those programs that affect a school as a community; thus, they are considered to be school-level interventions. Randomization of schools is the preferred method of choice when evaluating school-wide interventions; however, studies using this design are very costly because of the large number of schools generally required to provide the statistical power (precision) needed to obtain meaningful results. The purpose of this report is to demonstrate how prestudy archival school data combined with a stratification approach to analysis can be used to reduce the number of schools that are required in studies involving comprehensive, school-wide interventions. Obtaining baseline information from school archival data for stratification purposes prior to initiating a study assists researchers, school districts, and other stakeholders in choosing study sites (prior to randomization) that will provide the greatest likelihood that changes in behavioral outcomes can be measured if they occur.