Bayesian SEM PreviousNext
Mplus Discussion > Confirmatory Factor Analysis >
 Matthias Ziegler posted on Monday, April 25, 2016 - 6:22 am
Dear Linda,

I am trying to run a BSEM and keep getting an error message. The BSEM contains 5 latent variables, each has between 7 and 10 indicators. I want to assume cross-loadings and correlated residuals with priors as suggested in the white paper. However, regarding the correlated residuals, I do not want to allow them for the indicators from one latent variable. I keep getting the following error: THE PRIOR SPECIFICATION FOR PARAMETER 253 IS NOT AVAILABLE.
This parameter is the first residual covariance. If we allow all residual covariances, the model converges. However, this is not what we wanted to do. Do you have any advice?
 Bengt O. Muthen posted on Monday, April 25, 2016 - 6:30 pm
I don't think we can do this.
 Leonhard A. Bakker posted on Wednesday, June 29, 2016 - 4:59 am
Dear Bengt and Linda,
How would you classify a PPP-value of 0.233?
Kind regards, Leonhard
 Linda K. Muthen posted on Wednesday, June 29, 2016 - 6:51 am
That reflects decent fit.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message