Multilevel Latent Growth w/ Unstructu... PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
Message/Author
 James L. Lewis posted on Thursday, October 25, 2007 - 9:53 pm
I know how to do multilevel growth modeling with time unstructured data and an "observed" dependent variable. I have a new dataset, however, where the DV is made up of several observed variables at each time point (the same observed variables at each point), is time unstructured and has 2 levels (3 if you count the occasions nested in individuals). I have a feeling that it is right under my nose, but I don't know "the name" of this analysis and where to start reading about it. I imagine it is "Latent Growth Curve Analysis" but I dont know where to look if I want understand the time unstructured and multilevel aspects. Any suggested readings or can anyone tell me the particular "name of this?" Thanks a bunch.
 Linda K. Muthen posted on Friday, October 26, 2007 - 9:00 am
It sounds like you may have a multiple indicator growth model where development is over the factors. See if Example 9.15 is what you mean.
 James L. Lewis posted on Sunday, October 28, 2007 - 11:17 am
Yes, after reading example 9.15. This seems correct - a multiple indicator growth model. But to be sure allow me to describe how it is that my "Time" variable is (un)structured. The nature of the individually-varying times of observations is something like this:

ID Months
001 5
001 13
001 18
002 7
002 15
003 3
003 12
003 22
004 16
...

And NOT like this:

ID Months
001 5
001 15
001 25
002 5
002 15
003 5
003 15
003 25
004 15
...

Will this still work?? In most chapters on latent change analysis I have read, one way to test "slopes" is to respectively fix the parameters going from the latent "slope" variable to its indicators to 0,1,2,3 (etc). But with the time unstructured nature of my data this seems inappropriate (?).
Thanks a bunch.

JL
 Linda K. Muthen posted on Sunday, October 28, 2007 - 2:18 pm
If you have individually-varying times of observation, you would specify the model using the AT and TSCORES options as shown in Example 6.12.
 James L. Lewis posted on Sunday, November 04, 2007 - 8:16 pm
Thank you very much. This sent me down the path I was looking for.
J
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: