Search

Methodological challenges associated with structural equation modeling (SEM) and structured means modeling (SMM) in research on school violence and related topics in the social and behavioral sciences are examined. Problems associated with multiyear implementations of large-scale surveys are discussed. Complex sample designs, part of any large-scale survey, introduce multiple methodological challenges into SEM analysis. Problems with missing data and data imputation procedures are explored, along with commonly occurring departures from normality. Hypothesis testing of models and associated fit indices are explicated. Issues surrounding the use of Lagrange Multiplier tests of constraint validity in SMM are examined. Finally, basic problems associated with developing viable models of complex social phenomena with SEM are considered. [ABSTRACT FROM AUTHOR]