ABSTRACT
In previous chapters we learned about the basics of structural equation modeling. In this chapter we consider SEM models that demonstrate the variety of applications suitable for structural equation modeling. You should be aware, however, that our discussion will only introduce these SEM models. You are encouraged to explore other examples and applications reported in books (Marcoulides & Schumacker, 1996, 2001), SEM software programs (Amos, EQS, and LISREL), and the other references
13.1 MULTIPLE INDICATORS AND MULTIPLE CAUSES (MIMIC) MODELS
The term MIMIC refers to multiple indicators and multiple causes and defines a particular type of SEM model. The MIMIC model involves using latent variables that are predicted by observed variables. We illustrate by using an example from Jöreskog and Sörbom (1996a, Example 5.4, pp. 185-187) in which a latent variable (social participation) is defined by church attendance, memberships, and friends seen, This latent variable is predicted by the observed variables, income, occupation, and education. The MIMIC model is diagrammed in Fig. 13.1.
