Complex Sampling Design and Subpopula... PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
Message/Author
 Jay Unick posted on Friday, March 10, 2006 - 6:03 pm
Hi,

I have a question regarding LCA and complex sampling designs. I am using the National Comorbidity Survey that includes strata, psu’s and weights. But, I am only using a sub-sample of that survey not the whole population for my analysis, however, all the PSUs and Strata are present. I have been warned that when using a program such as Stata I need to have the whole population present but use a command subpop to subset my data in order to obtain accurate estimates. My question is, do I need to let Mplus 3.13 know that I am using a subpopulation and if so how can I do that?

Jay
 Bengt O. Muthen posted on Friday, March 10, 2006 - 6:28 pm
You need Mplus Version 4 to handle subpopulations with complex survey data - see the Version 4 User's Guide on our web site, page 403, which includes a reference to Korn & Graubard's book.
 anonymous posted on Wednesday, March 31, 2010 - 8:38 am
Hi,
I have used the subpopulation command so that only a subgroup of participants are analyzed in an LCA. However, when I use the subpopulation command in conjunction with assigning participants to classes(using the savedata command), all participants are assigned, even though they were not included in the LCA. Is this correct?
thanks!
 Linda K. Muthen posted on Thursday, April 01, 2010 - 10:30 am
This is correct. The way subpopulation is implemented is that we set to 0 the sampling weights for people that are not in the subpopulation. With zero weight they do not contribute to the estimation, however, the SAVEDATA command which is simply a data saving tool will include the people that are not in the subpopulation.
 Keng-Han Lin posted on Monday, May 10, 2010 - 9:14 am
Hi,

I'm using national survey data.
When I save the assigned classes by using SAVEDATA command, it seems that the number of people in each class doesn't match what I got from "CLASSIFICATION OF INDIVIDUALS BASED ON THEIR MOST LIKELY LATENT CLASS MEMBERSHIP" in the output.
I also gave it a try, the number matches when I took away settings of survey design.

How do I save the correct assigned classes given survey design?
Thanks!
 Linda K. Muthen posted on Monday, May 10, 2010 - 9:27 am
When a weight variable is used, the saved posterior probabilities are not weighted. I believe this is the difference you are seeing. If not, please send your files and license number to support@statmodel.com.
 Keng-Han Lin posted on Monday, May 10, 2010 - 10:00 am
Hi Linda,

Yes, I think this explains why there's a discrepancy.
But, is there a way to save weighted posterior probabilities (assigned classes)? or there's a reason we shouldn't do this?
Thank you.
 Linda K. Muthen posted on Tuesday, May 11, 2010 - 9:46 am
There is not a way to save weighted posterior probabilities. You would need to use weights if you use them.
 Keng-Han Lin posted on Tuesday, May 11, 2010 - 11:49 am
Hi Linda,

Then, how do we determine the assigned classes for each participant given weight?
I'm not quite sure how to calculate weighted posterior probabilities for each person.
Could you please provide the formula or reference that I can use?
Thank you.

Keng-Han
 Linda K. Muthen posted on Wednesday, May 12, 2010 - 8:26 am
You need to multiply the posterior probabilities by the weight variable.
 Den Shin Li posted on Wednesday, August 29, 2012 - 3:42 pm
Dr. Muthen,

I am new with complex survey design, and I need your advice on this.
I am running multiple group CFA with a national dataset, I kept receiving the following error message:

"*** ERROR
Cluster variable RES has multiple uses."


NAMES ARE
sex arg bul stub dis res get feel conf rel;
USEVARIABLES ARE arg bul stub dis res get feel conf;
grouping is rel(1=no 2=yes);
cluster IS rel;

ANALYSIS:
Type=complex;
ESTIMATOR IS MLR;

What should I do?

Thank you
 Den Shin Li posted on Wednesday, August 29, 2012 - 3:47 pm
Sorry, I need to make correction to my previous message with the problem I encountered:

"*** ERROR
Cluster variable REL has multiple uses."


NAMES ARE
sex arg bul stub dis res get feel conf rel;
USEVARIABLES ARE arg bul stub dis res get feel conf;
grouping is rel(1=no 2=yes);
cluster IS rel;

ANALYSIS:
Type=complex;
ESTIMATOR IS MLR;
 Linda K. Muthen posted on Wednesday, August 29, 2012 - 4:05 pm
You can't use REL as a grouping variable and a cluster variable. It can be used for only one function.
 Den Shin Li posted on Wednesday, August 29, 2012 - 5:45 pm
Dr. Muthen,

I apologize for bothering you again. When I removed grouping is REL, another error message appears in Model command:

"Unknown group name NO specified in group-specific MODEL command."

How can I specified groups while using cluster for complex survey design for MCFA?

Thanks.
 Linda K. Muthen posted on Thursday, August 30, 2012 - 10:27 am
Please send your output and license number to support@statmodel.com.
 Cecily Na posted on Friday, August 07, 2015 - 7:53 am
Hello Drs.,
I am selecting 10 strata out of 15 to do my analysis because these 10 strata are more similar to each other and are the focus of my research question.
When I specify the subpopulation, weight, strata, type=complex, my results should be generalizable to the populaiton in these 10 strata, right? I want to make sure my code is correct, and I'm not modeling for the whole population.
Thank you!
 Cecily Na posted on Friday, August 07, 2015 - 3:52 pm
Hello Drs.,
I am selecting 10 strata out of 15 to do my analysis because these 10 strata are more similar to each other and are the focus of my research question.
When I specify the subpopulation, weight, strata, type=complex, my results should be generalizable to the populaiton in these 10 strata, right? I want to make sure my code is correct, and I'm not modeling for the whole population.
Thank you!
 Cecily Na posted on Friday, August 07, 2015 - 4:02 pm
Hello Drs.,
I am selecting 10 strata out of 15 to do my analysis because these 10 strata are more similar to each other and are the focus of my research question.
When I specify the subpopulation, weight, strata, type=complex, my results should be generalizable to the populaiton in these 10 strata, right? I want to make sure my code is correct, and I'm not modeling for the whole population.
Thank you!
 Bengt O. Muthen posted on Friday, August 07, 2015 - 4:59 pm
Should be ok. The weights hold for all strata.
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: