Obtaining means in LGMM with covariates PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
 Anthony Mancini posted on Wednesday, April 15, 2009 - 11:57 am
I have a 4-class LGMM in which I am regressing class membership and growth factors on covariates.

After I introduced the regression of the growth factors into the model, I have had considerable difficulty figuring out how to generate the estimated class means.

I realize that there is no Means table anymore and that the intercept values now represent the class specific means based on the regressions of the growth factors.

What I have been able to figure out is how to generate the means that are reproduced in the Figure of the trajectories. An earlier post said that if you use grandmean centering this will subtract the sample mean and give you the correct means. Although this approximated the means in the figure, it was still not accurate.

I want to produce the mean values for a table and be sure they match the values in the Figure. Any help is greatly appreciated.
 Linda K. Muthen posted on Friday, April 17, 2009 - 10:42 am
You can obtain the means by asking for TECH4 in the OUTPUT command.
 Anthony Mancini posted on Wednesday, April 22, 2009 - 7:39 am
Ok, thank you. This did provide the means that matched the figures.

But TECH4 does not provide the standard errors or z scores for the class-specific growth factors.

Is it possible to get (or derive) the SEs and z scores as well so that they can be reported in the paper as well.

Thanks again in advance.
 Linda K. Muthen posted on Wednesday, April 22, 2009 - 9:37 am
If you want the standard errors, you would need to use MODEL CONSTRAINT to define the means. See the user's guide for more information about MODEL CONSTRAINT.
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