Effectivity intervention:comparison b... PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
Message/Author
 Carlijn C posted on Monday, June 23, 2014 - 1:42 pm
I'm studying the effectiveness of a intervention (with a RCT).
First, I study the normative development by a separate analysis of the control group. The four growth parameters (intercept/slope mean and variance) found in the model of the control group are repeated in the model for the experimental group, and I test equality of parameters.
I've used this syntax to test the difference in intercept variance:
model: i s | BVLT1@0 BVLT2@1 BVLT3@2;
[i@2.447]; [s@-0.086]; i@0.172 ; s@0.006 ;
Model control: i s | BVLT1@0 BVLT2@1 BVLT3@2;
[i@2.447] (p1); [s@-0.086] (p2); i@0.172 (p3); s@0.006 (p4);
Model experimental: i s | BVLT1@0 BVLT2@1 BVLT3@2;
[i] (p10); [s@-0.086] (p2); i@0.172 (p3); s@0.006 (p4);
Model test:
p1=p10;
But this doesn't work, I got an error.
 Carlijn C posted on Monday, June 23, 2014 - 1:44 pm
I've also used another syntax:
model: i s | BVLT1@0 BVLT2@1 BVLT3@2;
[i] (p1); [s] (p2); i (p3); s (p4);
Model control: i s | BVLT1@0 BVLT2@1 BVLT3@2;
[i] (p1); [s] (p2); i (p3); s (p4);
Model experimental: i s | BVLT1@0 BVLT2@1 BVLT3@2;
[i] (p1); [s] (p2); i (p3); s (p4);
Model test;
p1=p10;

Then I get also an error because this model doesn't fit anymore (slope variance is negative).
So now I'm using two different models for the control group (i s | BVLT1@0 BVLT2@1 BVLT3@2) and experimental group (i s | BVLT1@0 BVLT2@1 BVLT3@2; s@0). Is this the right way?
 Linda K. Muthen posted on Monday, June 23, 2014 - 4:09 pm
Please send the outputs showing the errors and your license number to support@statmodel.com.

Please limit your posts to one window.
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: