Structural Equation Modeling

Structural Equation Modeling (SEM) is a largely confirmatory, rather than exploratory, technique. It refers to a diverse set of mathematical models, computer algorithms, and statistical methods that fit networks of constructs to data. It is also is a very general, chiefly linear, chiefly cross-sectional statistical modeling technique. Structural Equation Modeling is commonly justified in the social sciences because of its ability to impute relationships between unobserved constructs from observable variables. In other hand SEM methods generally lack widely accepted goodness-of-fit statistics and most SEM software offers little latitude for error analysis.