Thank you for providing such a great resource, and for your consideration.
I have 2 models involving latent variables A, B, and C (and single data set). In Model 1 latent variable A predicts latent variable B, and in in Model 2, latent variable A predicts latent variable B and also latent variable C (B and C are mental health symptom clusters that are highly correlated). So, both models include variable A as the only predictor. Should the coefficient between path A and B remain the same in these 2 models, and if so, why?
In examining the Estimated Covariance Matrix for the Latent Variable, I noticed that the outcome variable in Model 1 is slightly reduced in Model 2 (the model that adds a second outcome latent variable). Why would this occur? There is no missing data and both models are of equal number of responses for all items. I am trying to understand what specifically changes when adding an outcome variable to a model.