Doubly-latent with latent interaction... PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
Message/Author
 Marcus Pietsch posted on Tuesday, May 26, 2020 - 6:42 am
Hello,

I try to model a twolevel model that includes 3 latent variable on both levels, a latent interaction on both levels as well as a dummy variable on both levels.

I modelled the dummy on both levels and centered it at level 1:

deccm = cluster_mean(devcom);

center devcom (groupmean);

Due to the latent interaction I need to add a two level random model:

Type = twolevel random;

Due to the dummy variable I need an integration algorithm:

Algorithm = Integration;

It is not possible to use the standard integration function, thus i apply montecarlo:

Integration=montecarlo(500);

As I need to use a random model, I am unable to use the WLSMV estimator.

Because of this the estimation takes many hours. Is there a way to speed up the estimations?

Thanks.
 Bengt O. Muthen posted on Tuesday, May 26, 2020 - 10:32 am
Check our FAQ: TECH8 negative ABS changes

If this doesn't help, we need to see your full output of your slow run including TECH8 to be able to diagnose. If it doesn't converge, just set the iterations to a smaller number to show where it is heading. Send to Support along with your license number.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: