LGC Model Non-Convergence PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
 Rebecca M Sacks posted on Thursday, May 28, 2015 - 12:32 pm
Thanks in advance for any insight people have to offer! I am in the very early stages of fitting a four time point model where the data exhibit non-linearity. I specified a freed loading growth model where time point 1 is fixed at 0, time point 2 is fixed at 1 and time points 3 and 4 are free, and am fitting the model separately for boys and girls.

The model in which I simply fit the growth curves ran fine. When I attempted to regress latent intercept and slope of the trajectory on an observed binary variable, however, the model failed to converge for boys (it was fine for girls). It will converge if I fix the loadings for the dependent variables. Is that an appropriate solution given the non-linearity of the trajectory? Any other thoughts?

Thank you!
 Bengt O. Muthen posted on Thursday, May 28, 2015 - 2:03 pm
Don't do that - instead explore the reason for the non-convergence.
 Rebecca M Sacks posted on Thursday, May 28, 2015 - 2:19 pm
Alright. I think I understand what you mean. As suggested in the User's Guide I looked at some of the diagnostic outputs and changed my starting values. It seems to have worked. Thank you.
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