This research provides a novel method for discovering group-level differences on endogenous variables in a multilevel structural equation modeling context. Furthermore, methods for calculating associated significance values of these group-level differences is described. This builds on current techniques for discovering group-level differences in multilevel regression models by extending this capability to full multilevel structural equation models. The included analysis provides a verification mechanism for the proposed method in a multilevel regression context with other current software. This provides verification that the method can then be extended to multilevel structural equation models.
Luse, Andy, "Estimating Random Effects in Multilevel Structural Equations Models" (2014). MWAIS 2014 Proceedings. 13.