Generating a new latent variable that... PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
Message/Author
 Shige Song posted on Wednesday, December 20, 2006 - 9:19 pm
the difference of another two latent variables. For example, I have two sequential growth processes and I want to get the difference of the two intercepts:

----------------
level trend | ht0*-12 ht1*-11 ht2*-10 ht3*-9 ht4*-8 ht5*-7 ht6*-6 ht7*-5 ht8*-4 ht9*-3 ht10*-2 ht11@-1 ht12@0;

i s | ht91@0 ht94@1 ht98*2;

dif | i - level;
----------------

where "level" and "trend" are latent growth factors for the first process and "i" and "s" are latent growth factors for the second process. I want to create a new latent variable that is the difference of the two latent growth factors "level" and "i".

Mplus reports that this is a syntax error "Unable to expand variable list: I - LEVEL"

The interaction of two latent variables can be generated using the "xwith" statement, but I could not find how to get the difference between two latent variable .

Thanks for any suggestions!

Shige
 Bengt O. Muthen posted on Thursday, December 21, 2006 - 4:01 pm
You have to define a new factor that is equal to i - level. You can do that as follows:

fdif by;
fdif on i@1 level@-1; fdif@0; [fdif@0];

This defines the new factor fdif as

fdif = 0 + 1*i-1*level+0
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: