ABSTRACT
Numerous structural equation modeling (SEM) programs are available, all of which are capable of analyzing everything from simple path models through latent variable structural equation models. Degrees of freedom are an index of the parsimony of a path model: the more degrees of freedom, the more values constrained prior to estimation and thus the greater the parsimony. The results of simple path analyses are identical using SEM or multiple regression analysis, but SEM programs can analyze more complex models and have real advantages when analyzing overidentified models. Amos uses a graphic approach and is probably the most intuitive and easiest SEM program to use. Most SEM programs, including Amos, can use the correlation matrix and standard deviations as input for the analysis. SEM programs will produce tables of direct, indirect, and total effects for both the standardized and unstandardized solution.
