Effects of Students and School Factors on Five Measures of School Disorder

The research was based on Philadelphia middle school data collected by Welsh et al. (1997) during the 1994-95 school year. Several independent measures were taken from the 118-item Effective School Battery (ESB), whose reliabilities and validities have been established across diverse subgroups and settings. The five measures of disorder examined were victimization, safety, avoidance, offending, and misconduct. The study examined whether schools varied significantly on these five measures. If schools failed to differ on the dependent measures, there would be no variance to explain. The study also examined to what degree student (within-school) and between-school factors explained variance in different measures of school disorder. Finally, the study examined to what degree specific individual-level and school-level factors predicted different measures of school disorder. Survey responses were analyzed from 4,640 middle school students by using MANCOVA. Schools varied significantly on all five measures of disorder; both student characteristics and school climate variables provided significant explanatory power for each. Patterns of results varied, however, for different measures of disorder. Between-school effects were much stronger for students' misconduct than for more serious offending. This paper concludes with a discussion of implications for research and policy on school disorder. 4 tables and 88 references