Constraints in a GMM PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
Message/Author
 Darren L Dahly posted on Monday, August 31, 2009 - 6:53 am
I want to estimate a GMM based on a LGCM with freely estimated factor loadings and autocorrelated residuals. If I extend this model to a mixture model with 2 classes, and only specify an %overall% model, I have no problems. However, as soon as I try to free any additional parameters across classes (such as the factor loadings), I get the following error:


*** ERROR in MODEL CONSTRAINT command
A parameter label or the constant 0 must appear on the left-hand side
of a MODEL CONSTRAINT statement. Problem with the following:
BMI258* =

Ideally I'd like to look at a mixture model where every parameter if freely estimated between classes. Any ideas on where I'm going wrong? Many thanks.

bmi0-bmi24 (resvar1);
bmi0-bmi22 PWITH bmi2-bmi24 (p1);
bmi0-bmi20 PWITH bmi4-bmi24 (p2);
bmi0-bmi18 PWITH bmi6-bmi24 (p3);

bmi138-bmi258 (resvar2);
bmi138-bmi226 PWITH bmi192-bmi258 (p4);
bmi138-bmi192 PWITH bmi226-bmi258 (p5);
bmi138 PWITH bmi258 (p6);

MODEL CONSTRAINT:
NEW (corr1);
p1 = resvar1*corr1;
p2 = resvar1*corr1**2;
p3 = resvar1*corr1**3;

MODEL CONSTRAINT:
NEW (corr2);
p4 = resvar2*corr2;
p5 = resvar2*corr2**2;
p6 = resvar2*corr2**3;
 Linda K. Muthen posted on Monday, August 31, 2009 - 7:48 am
Please send your output and license number to support@statmodel.com.
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: