Structural Equations with Latent Variables.

Abstract

Within the past decade the vocabulary of quantitative research in the social sciences has changed. Terms such as "LISREL," "covariance structures," "latent variables," "multiple indicators," and "path models" are commonplace. The structural equation models that lie behind these terms are a powerful generalization of earlier statistical approaches. They are changing researchers' perspectives on statistical modeling and building bridges between the way social scientists think substantively and the way they analyze data. We can view these models in several ways. They are regression equations with less restrictive assumptions that allow measurement error in the explanatory as well as the dependent variables. They consist of factor analyses that permit direct and indirect effects between factors. They routinely include multiple indicators and latent variables. In brief, these models encompass and extend regression, econometric, and factor analysis procedures.