This analysis provided one strategy for analyzing longitudinal models containing a larger number of constructs and time points. A large number of parameters may tax the computational capacities of computers; however, advances assure faster com- putational speeds and higher RAM that can handle the potentially large numbers of simultaneous equations. Additionally, the diagrammer features in EQS make com- plex model building much simpler. Because of the large number of estimated pa- rameters, model convergence requires running smaller models that generate ap- propriate parameter start values in the full model. With a little ingenuity through use of nested models, it is possible to obtain relatively reasonable start values that could aid in the convergence of large models.