MG Growth Mixture Model (Syntax Issue... PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
Message/Author
 KK posted on Tuesday, March 20, 2007 - 3:04 pm
In a multiple group growth mixture modeling context, what is the effect of the input term (I've provided example code below):
c#1 on SEX#1 Ctemp;
where SEX is the grouping factor.
My understanding is that different parameters will be estimated for the latent classes as a function of sex, with also some effect of Ctemp (a continuous predictor). However, isn't this done directly with explicitly allowing differences in the parameters (as I've done below)?

Along similar lines, what does the following mean conceptually in the output (which is related to the input I asked about above):

Categorical Latent Variables

C#1 ON SEX#1
0.345 0.266 1.299

C#1 ON CTEMP
1.003 0.259 3.874

Means
SEX#1 0.067 0.056 1.202
C#1 -0.656 0.169 -3.886

I understand everything in the output and how to set-up the model except for these issues. Since multiple group growth mixture models are few and far between, I've not been able to find much on the issue.
Thanks!
 KK posted on Tuesday, March 20, 2007 - 3:05 pm
The post was too big so I had to split it. The syntax used is given below.

! BEGIN EXAMPLE CODE:
USEVARIABLES ARE p1 p2 p3 p4 p5 Ctemp;
IDVARIABLE = id;
MISSING IS *;
CLASSES = SEX(2) c(2);
KNOWNCLASS = SEX(csex_m01=1 csex_m01=2);

ANALYSIS:
TYPE IS mixture meanstructure missing;
starts=500 50;

MODEL:
%overall%
i s | p0@0 p1@1 p2@2 p3@3 p4@4;
i on Ctemp;
s on Ctemp;

c#1 on SEX#1 Ctemp;

%SEX#1.c#1% ! Male for class 1.
i with s;
[i*-.08 s*.71];
p1;
p2;
p3;
p4;
p5;
i on Ctemp;
s on Ctemp;

%SEX#1.c#2% ! Male for class 2.
i with s;
[i*-.76 s*-.09];
p1;
p2;
p3;
p4;
p5;
i on Ctemp;
s on Ctemp;

%SEX#2.c#1% ! Female for class 1.
i with s;
[i*-.52 s*-.83];
p1;
p2;
p3;
p4;
p5;
i on Ctemp;
s on Ctemp;

%SEX#2.c#2% ! Female for class 2.
i with s;
[i*.06 s*.49];
p1;
p2;
p3;
p4;
p5;
i on Ctemp;
s on Ctemp;
 Linda K. Muthen posted on Tuesday, March 20, 2007 - 9:32 pm
If your post is too long to fit, then it is too long for Mplus Discussion. Please do not double post in the future.

c#1 ON sex#1 means the the class probabilities of c vary across gender.
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: