Dropping out of Middle School: A Multilevel Analysis of Students and Schools

Prior research on dropouts has often focused on high schools and examined the issue from either the individual perspective or the institutional perspective. Using data from the National Educational Longitudinal Survey of 1988 and a new form of hierarchical linear modeling (HLM), this study focuses on dropouts from middle school and examines the issue from both individual and institutional perspectives. At the individual level, the results identified a number of family and school experience factors that influence the decision to leave school, with grade retention being the single most powerful predictor. But disaggregating the analysis also revealed that there are widespread differences in the effects of these factors on White, Black, and Hispanic students. At the institutional level, the results revealed that mean dropout rates vary widely between schools and that most of the variation can be explained by differences in the background characteristics of students. But restricting the analysis to lower SES schools shows widespread differences in both mean dropout rates and social class differentiation among such schools. Moreover, much of the variation among those schools can be explained by social composition of students and by several structural features of schools and school climate.