Cluster Variance PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
 mdehne posted on Wednesday, May 13, 2020 - 5:15 am
Dear Profs Muthéns:

I am conducting a two-level longitudinal multilevel analysis (within-person, between-person). I am confused due to the new error message that was implemented starting with Mplus version 8. For my dependent time-varying variable, I understand the error message for three clusters that show no variance across time. The model works fine for the independent predictor time variable(s)—linear/quadratic (l: 0,1,2,3,4 / q: 0,1,4,9,16).

To compute time-averaged mean slopes (see Mayer, Steyer, & Mueller, 2012), I used a orthogonal polynomial contrast coding (l: -2,1,0,1,2 / q: 2,-1,-2,-1,2). I cannot explain why the same error message as for the dependent variable, then, occurs for some but not all clusters for the quadratic orthogonal polynomial time predictor. I checked the data file: The time codings (as a transformation in SPSS) were carried correctly—the missing values as well. Could you please provide any advice?

The best comes at the end: If I add a predictor for my slopes, the error message disappears for some clusters fot the qudratic time predictor.

Thank you very much in advance.

Kind regards,
 Tihomir Asparouhov posted on Thursday, May 14, 2020 - 4:19 pm
I am not sure what error message you are referring to but it sounds like the missing time coding could be the problem. There is no reason to have missing values for the time. Send you example to for further help.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message