I am trying to test an SEM model that includes an interaction between continuous variables. My data are somewhat skewed, so I am looking for a set of analysis commands that will successfully address non-normality and allow me to use the XWITH command. Is this possible?
You can use the MLR estimator which is robust to non-normality. XWITH is for interactions involving one or more latent variables. If your variables are observed, you can use the DEFINE command to create the interaction.