Multiple group analysis in Monte Carl... PreviousNext
Mplus Discussion > Categorical Data Modeling >
Message/Author
 Siddhi Pittayachawan posted on Friday, March 12, 2010 - 2:49 am
Hi Muthen,

I am conducting multiple group analysis (i.e. 2 groups) and would like to add Monte Carlo as part of my study. The empirical model was assessed with WLSMV estimation. The ESTIMATES command was used to produce a population data file for the Monte Carlo study. However, when I ran the MC study, it said that "The POPULATION file contains too much data. Check your data."

I am quite new to the MC study, and I am not sure what went wrong. Could you please suggest me?

Regards,

Siddhi
 Linda K. Muthen posted on Friday, March 12, 2010 - 5:53 am
You need to specify you population parameter values using MODEL POPULATION rather than the ESTIMATES option.
 Siddhi Pittayachawan posted on Friday, March 12, 2010 - 2:00 pm
You mean that I simply record the estimated value in the empirical model and fix into the MODEL POPULATION in the Monte Carlo study?
 Linda K. Muthen posted on Friday, March 12, 2010 - 2:22 pm
Yes.
 Siddhi Pittayachawan posted on Friday, March 12, 2010 - 5:51 pm
Then what is the purpose of using the ESTIMATES command to save the population information from the empirical model? I thought that it contains the population values. Could you please elaborate this point?
 Linda K. Muthen posted on Friday, March 12, 2010 - 5:59 pm
The ESTIMATES option does not work for all models. If it works, then you don't have to enter the values. In your case, you must enter the values.
 Siddhi Pittayachawan posted on Friday, March 12, 2010 - 7:28 pm
Thank you very much for your explanation. Now I understand why it never worked.

So I guess that I should not use MODEL POPULATION but MODEL POPULATION-label command to specify each group, am I correct?

What about MODEL COVERAGE command? In the manual, it seems that they do the same thing. I searched on the board here, but I could not find any topic on this command at all.
 Linda K. Muthen posted on Saturday, March 13, 2010 - 6:35 am
Yes you should use MODEL POPULATION and MODEL POPULATION label to specify values for each group.

In Monte Carlo simulation studies, the MODEL command describes not only the analysis model but also provides values for each parameter that are used both as the population parameter values for computing coverage and as starting values in the estimation of the model. If you want to use the values only as starting values, you can use MODEL COVERAGE to give other values to be used for computing coverage.
 Siddhi Pittayachawan posted on Monday, March 15, 2010 - 1:17 am
Thank you very much, Linda. In the past few days, I have been experimenting Monte Carlo studies, although the setting up part is a bit tricky, and I found how it can be used in my research.
 Linda K. Muthen posted on Monday, March 15, 2010 - 6:20 am
Each of the user's guide examples has a Monte Carlo counterpart that is used to generate the data for the example. If you find an example that is close to your model, these can be a good place to start.
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: