EXAMPLE 1.  The nonlinear SEM example 1 shown in Wall (2009) is nonlinear 
in parameters and nonlinear in endogenous variables and cannot be fit using Mplus,
 but can be fit using Proc NLMIXED (maximum likelihood) and Winbugs (Bayesian).  

<UL><LI> <A HREF = "example1 - verynonlinear/verynonlinear (generate data and program).sas">
 SAS Proc NLMIXED code </A> - Besides fitting the nonlinear SEM, this program also generates the data.
<LI> <A HREF = "example1 - verynonlinear/nonlinear SEM.odc">Winbugs code</A>
</UL>

EXAMPLE 2. The interaction SEM example 2 shown in Wall (2009) can be fit using 
Mplus which uses maximum likelihood in addition to Proc NLMIXED (maximum likelihood) and Winbugs (Bayesian).

<UL> <LI> <A HREF = "example2 - interaction/generate data.sas">Code to generate the data</A> - A SAS program to generate the data
<LI> <A HREF = "example2 - interaction/interactionmodelmplus.inp">Mplus code</A>
<LI> <A HREF = "example2 - interaction/interactionmodel.sas">
 SAS Proc NLMIXED code </A> 
<LI> <A HREF = "example2 - interaction/interactionmodel.odc">Winbugs code</A>
</UL>