Properties of goodness of fit statistics for structural equation models

The subject of this thesis is goodness of fit statistics for structural equations models (SEMs). The main issue is to give the researchers who apply SEMs a coherent view on the properties of existing fit measures. We will show how the fir statistics can give new interesting conclusions about models and data sets. A nested models design will be proposed to show how the goodness of fit statistics can be used to choose the "statistically optimal" model among many competing alternatice models.