Covariances with multiple time varyin... PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
 mike parent posted on Thursday, December 12, 2013 - 2:46 pm
I'm building a model but I wasn't able to find an example in the manual or handouts with multiple time-varying covariates. I wanted to make sure I have the covariances set up correctly.

I have
5 waves
1 time-invariant covariant (TIC)
3 time-varying covariants (TVC1-TVC3) collected at all 5 waves
1 DV collected at all 5 waves

Based on Topic 3's handouts, I *think* I should be setting the covariances up so that I have covariances
1- within each TVC across time, and
2- across TVCs within time, and
3- every TVC with the TIC
But NOT across measure and time (e.g., not TVC1_time1 with TVC2_time4).

Is this a correct understanding of setting up the covariances in a model with more than one TVC, or am I going astray?

 Bengt O. Muthen posted on Friday, December 13, 2013 - 10:38 pm
All covariates are correlated as the default by Mplus (a particular TVC at one time point is considered a different covariate than the same TVC at a different time point). Just don't mention their parameters and this will occur (Mplus then treats them as "x" variables).
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