LCGM: A continuous and a count DV PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
Message/Author
 Michael Wade posted on Monday, December 21, 2015 - 8:52 am
Hi - I’m conducting a latent variable growth mixture model with two DVs measured over time. One of the DVs is continuous, whereas the other is a count variable. I’m having problems estimating random effects associated with the count variable, namely co-variances.
FYI: The “with” statements below produce an error:
“This analysis requires numerical integration for variable(s) I2. Add ALGORITHM=INTEGRATION to the ANALYSIS command.” My program will not run when I add “ALGORITHM=INTEGRATION” to the analysis command….?

VARIABLE: NAMES ARE id Bin1-Bin6 ptsd1-ptsd6;
usevar Bin1-Bin6 ptsd1-ptsd6 ;
CLASSES = C(2);
COUNT = Bin1-Bin6(p);
ANALYSIS: TYPE=MIXTURE MISSING ;
ESTIMATOR IS MLR;
MODEL: %OVERALL%
I1 S1| PTSD1@0 PTSD2@1 PTSD3@2 PTSD4@3 PTSD5@4 PTSD6@5;
I2 S2| Bin1@0 Bin2@1 Bin3@2 Bin4@3 Bin5@4 Bin6@5 ;
I1 with S1; I1 with I2; I1 with S2; I2 with S2; S1 with S2;
 Linda K. Muthen posted on Monday, December 21, 2015 - 12:26 pm
What do you mean by won't run. What happens.
 Michael Wade posted on Monday, December 21, 2015 - 2:49 pm
Well, the program runs for about 10seconds, then the cmd.exe (DOS) window closes. Nothing is written to the output file.

Does my syntax look correct?
 Linda K. Muthen posted on Monday, December 21, 2015 - 4:12 pm
Please send the input, data set, and your 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: