Latent Decomposition of Independent V... PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
Message/Author
 James Algina posted on Sunday, June 10, 2018 - 8:10 am
When I run this simple model using 8.1

analysis:
type=twolevel;
model:
%within%
Y on X;
%between%
Y on X;

I get this warning message:

In the MODEL command, the x variable on the WITHIN level refers to the whole observed X variable.

Is this warning message accurate?

Thanks

Jamie
 Bengt O. Muthen posted on Sunday, June 10, 2018 - 9:53 am
Please send the full output to Support along with your license number.
 Bengt O. Muthen posted on Monday, June 11, 2018 - 10:34 am
This warning message should be printed only with random slopes like for the UG ex 9.2c (part 3) on pages 278-279. So you should ignore this warning for your example. A latent variable decomposition is carried out also with ML (not only for Bayes) for your example and the results are correct.
 Michael Sciffer posted on Wednesday, June 20, 2018 - 1:20 pm
Hi Bengt

I am also now getting this error since the latest update to version 8.

I am trying to use two-level latent decomposition of an x covariate with a weighted dataset so my only option is MLR estimation.

So I can ignore this error?
 Bengt O. Muthen posted on Wednesday, June 20, 2018 - 2:22 pm
Yes.
 DONGHO CHOI posted on Monday, January 06, 2020 - 11:04 pm
Hello,

I both tried UG ex 9.1b and 9.2c and they both gave same error message above. Yet in 9.2c UG says the model uses observed covariate x on within-level and latent variable covariate x_bj on between-level.

So is warning message in 9.2c not ignorable? I mean are the x scores being used on within-model in 9.2c actually observed scores, not latent ones?

Thanks!
 Bengt O. Muthen posted on Tuesday, January 07, 2020 - 2:41 pm
Mplus version 8.4 gives the following message for ex 9.2c and it is correct:

In the MODEL command, the predictor variable on the WITHIN level refers to the whole
observed variable in the statement: S | Y ON X
To use the latent within-level part, use ESTIMATOR=BAYES in the ANALYSIS command.

If you have a specific output question, send the full output to Support along with your license number.
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: