Fixing residual variance to zero PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
Message/Author
 Sue Walker posted on Tuesday, March 12, 2013 - 9:09 pm
Hi

This is my model below. I have a low residual variance on ftconfs. How can I set this to 0?


i s | dtconfs@0 etconfs@1 ftconfs@2;
s (slope);
i with s@0;
%c#1%
[i*1.3];
%c#2%
[i*3.5];

MODEL CONSTRAINT:
slope > 0;
 Linda K. Muthen posted on Wednesday, March 13, 2013 - 8:14 am
If you want to fix s at zero, say in the MODEL command:

s@0;
 Sue Walker posted on Monday, March 25, 2013 - 10:32 pm
Thanks Linda, isn't it the residual variance on ftconfs that I want to set at zero?
 Linda K. Muthen posted on Tuesday, March 26, 2013 - 6:08 am
Then you would say

ftconfs@0;
 Sophie Potter posted on Wednesday, December 05, 2018 - 12:56 am
Hi,
My growth model wont run unless I set the residual variances to zero.

When I try to let them be estimated freely (see below) the model will not run. Do you know why?

ANALYSIS:
TYPE = RANDOM;
ESTIMATOR=MLR;
H1ITERATIONS=20000;
ITERATIONS=20000;
COVERAGE=0;
ALGORITHM=INTEGRATION;

MODEL:
!define level factors
i s q | WB12 AT TS0
WB14 AT TS1
WB16 AT TS2;


!means
[WB12-WB16@0];
[i*12.226];
[s*-0.199];
[q*0.115];

!variance and covariance
i*4.829;
s*8.856;
q*;
i WITH s*;
i WITH q*;
s WITH q*;

!specific variance
WB12*1.983;
WB14*2.584;
WB16*1.236;
 Bengt O. Muthen posted on Wednesday, December 05, 2018 - 1:28 pm
Send your data and output with the residual variances free to Support along with your license number. Include TECH8 in your run.
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: