What do you mean by "simplified approach"? Without latent variables? Easier to handle as other programs?
Within SPSS it is not possible to do SEM, but a (very) simplified approach is path analysis, without latent variables. Otherwise AMOS as SPSS additional package.
Mplus is also a good choice, it has a Limited free student version available. The advantage to other SEM software is that Mplus allows you to build models for various varible types: continuous, categorical, count, survival etc. It can also handle skewed distributions. Mplus interface is easier to start because it is not based on matrix speifications like Lisrel. See http://www.statmodel.com/demo.shtml for further information about the demo.