Setting error terms to zero? PreviousNext
Mplus Discussion > Structural Equation Modeling >
Message/Author
 pat berglund posted on Tuesday, April 10, 2007 - 10:22 am
Hello, I would like to know how to set observed x var error terms to zero for one model and then allow the error terms to be calculated as usual for the next model.

I have 5 obs. x vars and create a f1 vars as follows:
model:
f1 by x1-x5 ;
!run 2 models using f1 ;
sub on f1 ;
dep on f1 ;

How can I set the x1-x5 error terms to 0 for these models and then rerun with the 5 x vars allowed to calculated error terms as usual? Thanks, am new to Mplus and could use some assistance on this. Pat
 Linda K. Muthen posted on Tuesday, April 10, 2007 - 2:23 pm
The residual variances will be estimated in the model shown above if x1-x5 are continuous variables. See Chapter 16 for a description of all of the features of the MODEL command. To set the residual variances to zero, add

x1-x5@0;

to the MODEL command. This may result in a singular Sigma matrix. If so, you can use MODEL TEST.
 pat berglund posted on Tuesday, April 10, 2007 - 4:46 pm
Thanks very much, I will give this a try. Pat Berglund
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: