Multilevel LGC with unequal time inte... PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
Message/Author
 Eeske van Roekel posted on Thursday, January 10, 2013 - 8:20 am
Dear Dr. Muthen,

We have multilevel data (multiple assessments nested within participants) on which we want to conduct linear growth curve analyses. Our main goal is to examine how positive and negative affect develop after a certain (predefined) event. However, our assessments were randomly timed for each individual during the day, and hence are not spaced equally for everyone. For example, for individual A; T1 can be 30 minutes after the event, T2 at 64 minutes, T3 at 99 minutes etc., whereas for individual B; T1 can be at 56 minutes, T2 at 71 minutes, T3 at 143 minutes, etc. Therefore, we wanted to use the AT function, to control for the unequal distribution of time between measurement points. However, we cannot find in the manual or on the forum whether this is possible in a multilevel model. Importantly, we do not intend to include time-varying covariates, but we do want to examine whether level 1 (assessment level) and level 2 (individual level) predictors are related to the intercept and slope. Can you let us know whether this is possible and how we should model this?

Thanks for your help,
Eeske
 Linda K. Muthen posted on Thursday, January 10, 2013 - 4:12 pm
AT is available with TYPE=TWOLEVEL. Specify it on the within level. The growth factors will have parameters on both levels.
 Eeske van Roekel posted on Friday, January 11, 2013 - 7:09 am
Thanks for your reply. I have tried this, but I get a 'fatal error'. This is how I specified the model:
ANALYSIS: TYPE = TWOLEVEL RANDOM;
ALGORITHM=INTEGRATION;
MODEL:
%WITHIN%
iw sw qw| NA0-NA3 AT Timet0-Timet3;
NA0-NA3 (1);
%BETWEEN%
ib sb qb| NA0@0 NA1@1 NA2@2 NA3@3 ;
NA0-NA3@0;

The error I get is: FATAL ERROR
THERE IS NOT ENOUGH MEMORY SPACE TO RUN Mplus ON THE CURRENT INPUT FILE. THE ANALYSIS REQUIRES 8 DIMENSIONS OF INTEGRATION RESULTING IN A TOTAL OF 0.25629E+10 INTEGRATION POINTS. THIS MAY BE THE CAUSE OF THE MEMORY SHORTAGE. YOU CAN TRY TO FREE UP SOME MEMORY BY CLOSING OTHER APPLICATIONS THAT ARE CURRENTLY RUNNING. NOTE THAT THE MODEL MAY REQUIRE MORE MEMORY THAN ALLOWED BY THE OPERATING SYSTEM. REFER TO SYSTEM REQUIREMENTS AT www.statmodel.com FOR MORE INFORMATION ABOUT THIS LIMIT.

I hardly had any other applications running, so I doubt that would be the problem.
 Linda K. Muthen posted on Friday, January 11, 2013 - 5:00 pm
Try the following MODEL command:

MODEL:
%WITHIN%
i s q| NA0-NA3 AT Timet0-Timet3;
NA0-NA3 (1);
%BETWEEN%
NA0-NA3@0;

With AT you specify the model only on the within level. i, s, and q are available on both levels.
 Eeske van Roekel posted on Monday, January 14, 2013 - 10:46 am
Thank you for your reply. When I run this model, I still get the following error (I have tried this on two different computers, that meet the system requirements):

FATAL ERROR
THERE IS NOT ENOUGH MEMORY SPACE TO RUN THE PROGRAM ON THE CURRENT
INPUT FILE. THE ANALYSIS REQUIRES 5 DIMENSIONS OF INTEGRATION RESULTING
IN A TOTAL OF 0.75938E+06 INTEGRATION POINTS. THIS MAY BE THE CAUSE
OF THE MEMORY SHORTAGE. YOU CAN TRY TO FREE UP SOME MEMORY BY CLOSING
OTHER APPLICATIONS THAT ARE CURRENTLY RUNNING. NOTE THAT THE MODEL MAY
REQUIRE MORE MEMORY THAN ALLOWED BY THE OPERATING SYSTEM.
REFER TO SYSTEM REQUIREMENTS AT www.statmodel.com FOR MORE
INFORMATION ABOUT THIS LIMIT.

I read on the forum that you advised to add "INTEGRATION=MONTECARLO (5000);" to the analysis command in response to the same error. I have tried this, but then I get this:
THE ESTIMATED WITHIN COVARIANCE MATRIX COULD NOT BE INVERTED.
COMPUTATION COULD NOT BE COMPLETED IN ITERATION 2.
CHANGE YOUR MODEL AND/OR STARTING VALUES.
THE MODEL ESTIMATION DID NOT TERMINATE NORMALLY DUE TO AN ERROR IN THE
COMPUTATION. CHANGE YOUR MODEL AND/OR STARTING VALUES.
 Linda K. Muthen posted on Monday, January 14, 2013 - 12:06 pm
Please send your output and 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: