ABSTRACT
This chapter focuses on to link the path analytic framework to the techniques of structural modeling with latent variables; to introduce the reader to the Linear Structural Relationships (LISREL) program and to facilitate its use and interpretation. It determines the initiate but will provide a useful review for those somewhat more familiar with the approach and the LISREL program. The justification for using linear structural equation models with unmeasured variables, or LISREL, can be found in the sociological literature addressing measurement error in causal models. The model can estimate the parameters among the unobserved concepts of interest, derive the epistemic correlation between measured indicator and concept, and recognize and control for the effects of differential bias between indicators. The importance of using multiple indicator models, however, should be clear in the context of causal models and the ability of the LISREL program to facilitate a variety of such analyses should be evident.
