How Do We Make This Trapezable?

In these models, we are usually doing both the covariance structure and a mean structure. And when you're estimating a covarianant structure and a means structure simultaneously, historically that can present some estimation challenges. In fact, there's even a version of this type of model where you go in and you take all of your product terms and you subtract their means from them so that you wind up centering those. So that's a double mean centered strategy. The idea of that being that you can actually eliminate the mean structure of this model altogether and don't have to worry about that. Just allow all of the factors to co vary. Don't worry about constraining their variances. Don