Monte Carlo with Latent Profile Data
Message/Author
 Wenjuan Sang posted on Wednesday, December 03, 2014 - 10:48 am
Hello,
Sorry if this question is too basic. I am doing a latent profile analysis with 15 items (x1-x15), which is divided into 3 classes. As the sample size is smaller than 200, I want to do a Monte Carlo study to test the performance of the model. I went through the user guide and found example 7.3 should fit my case. However, I got a problem with changing the syntax.
The example is:
[c#1*1];
x1-x10*1;

%c#1%
[u1\$1*2 u2\$1*2 u3\$1*-2 u4\$1*-2];
[x1-x2*-2 x3-x4*-1 x5-x6*0 x7-x8*1 x9-x10*2];

%c#2%
[u1\$1*-2 u2\$1*-2 u3\$1*2 u4\$1*2];
[x1-x2*2 x3-x4*1 x5-x6*0 x7-x8*-1 x9-x10*-2];

It seems that the case is for an even number sample. Would someone please tell me how to make a change? Thank you!
 Bengt O. Muthen posted on Wednesday, December 03, 2014 - 4:39 pm
UG ex 7.9 would be more suitable given your continuous indicators.

You say you have 3 classes so you would have to give information about a 3rd class as well, not only 2 classes.

And I don't know what the u variables are in your application.
 Wenjuan Sang posted on Saturday, December 06, 2014 - 1:00 pm
Thank you so much!
I checked that example and wrote as following.
MONTECARLO:
NAMES = X1 X2 X3 X4 X5 X6 X7 X8 X9 X10 X11 X12 X13 X14 X15;
GENERATE=X1-X15(1);
NOBS = 167;
NREPS = 100;
SEED =12345;
GENCLASSES = c(3);
CLASSES = c (3);
NCLASSES = 3;
REPSAVE = ALL;
SAVE= C:\LPA.dat;
ANALYSIS: TYPE = MIXTURE;

model population:
%overall%
[c#1*0];

%c#1%
[X1-X15*1];
X1-X15*1;

%c#2%
[X1-X15*-1];
X1-X15*1;

%c#3%
[X1-X15*1];
X1-X15*1;

model
%overall%

[c#1*0];
X1-X15*1;

%c#1%
[X1-X15*1];

%c#2%
[X1-X15*-1];

%c#3%
[X1-X15*1];

OUTPUT: TECH8 TECH9;
Mplus showed that "The DATA command is not available in conjuction with the MONTECARLO command." I would be really appreciate if anyone could take a look of the syntax.
 Bengt O. Muthen posted on Saturday, December 06, 2014 - 1:54 pm