Two-level growth models for a categor... PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
Message/Author
 Hee-Jin Jun posted on Wednesday, September 21, 2011 - 1:41 pm
Dear Linda,

I am running two-level growth model for a categorical outcome. My predictors are time-varying. Below is my codes and I got an error message saying “SINGULAR INFORMATION MATRIX PROBLEM OCCURRED IN THE UNIVARIATE ESTIMATION OF VARIABLE AMTBNG97.” This error message was for all outcome variables up to AMTBNG05.
In fact, I don’t know how to fix these errors. Would you please help how to fix these errors?

Thanks a lot.

Hee-Jin

due to large size, I will add the program to the next post
 Hee-Jin Jun posted on Wednesday, September 21, 2011 - 1:45 pm
CATEGORICAL = amtbng97 amtbng98 amtbng99 amtbng00 amtbng01 amtbng03 amtbng05;
WITHIN = msthet97 msthet98 msthet99 msthet00 msthet01 msthet03 msthet05 bilgay97 bilgay98 bilgay99 bilgay00 bilgay01 bilgay03 bilgay05 sother97 sother98 sother99 sother00 sother01 sother03 sother05 gender;
BETWEEN = CLIM4;
CLUSTER = stnum;
CENTERING = GRANDMEAN (clim4);
ANALYSIS: TYPE = TWOLEVEL;
INTEGRATION = 7;
PROCESS = 4;
ESTIMATOR = WLSM;
MODEL: %WITHIN%
iw sw | amtbng97@0 amtbng98@1 amtbng99@2 amtbng00@3 amtbng01@4 amtbng03@6 amtbng05@8;
iw sw ON gender;
amtbng97 ON msthet97 bilgay97 sother97 ;
amtbng98 ON msthet98 bilgay98 sother98 ;
amtbng99 ON msthet99 bilgay99 sother99 ;
amtbng00 ON msthet00 bilgay00 sother00 ;
amtbng01 ON msthet01 bilgay01 sother01 ;
amtbng03 ON msthet03 bilgay03 sother03 ;
amtbng05 ON msthet05 bilgay05 sother05 ;
%BETWEEN%
ib sb | amtbng97@0 amtbng98@1 amtbng99@2 amtbng00@3 amtbng01@4 amtbng03@6 amtbng05@8;
ib sb ON CLIM4;
 Linda K. Muthen posted on Wednesday, September 21, 2011 - 2:33 pm
Please send the full output and your license number to support@statmodel.com. In the future, please limit your posts to one window.
 Katerina Gk posted on Thursday, October 10, 2013 - 11:04 am
Dear Linda,

I have already standardized my observed values and I have a twolevel model and I, also, aggregate the observed variables on the between level,using the command cluster_mean as you reccomend me in a previous message.

I get the same error SINGULAR INFORMATION MATRIX PROBLEM OCCURRED IN THE UNIVARIATE
ESTIMATION OF VARIABLE ...

SINGULAR INFORMATION MATRIX PROBLEM OCCURRED IN THE BIVARIATE ESTIMATION OF VARIABLES ...

Can you recommend me something to fix this error???

When I run the same model without standarized my values then it takes too long to converged anf give the fatal error, that there is not enough memory.

Thank you very much
 Linda K. Muthen posted on Friday, October 11, 2013 - 10:37 am
I am not sure why you have standardized your variables. This is only correct if you have a scale free model. Please send the output before you standardized the variables and your license number to support@statmodel.com.
 Katerina Gk posted on Saturday, October 12, 2013 - 9:41 am
Dear Linda,

thank you for your message, I ckecked again my data I decided not to standardized the variables. At last I refix my models and the models run.
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: