Twolevel random slopes: Bayesian mode... PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
Message/Author
 Tino Nsenene posted on Monday, September 26, 2016 - 9:30 am
Hello,

I am working on a twolevel SEM with random slopes / cross level interactions and Estimator = Bayes. To compare the relative fit of different models (e.g. model with vs. without random slopes, model with cross-level interaction vs, model with random slopes only) I use the DIC. Which additional fit measures availabel in Mplus would you recommend?

Best
Tino
 Bengt O. Muthen posted on Monday, September 26, 2016 - 4:03 pm
No other ones are available.
 Tino Nsenene posted on Wednesday, September 28, 2016 - 7:10 am
Thank your for you quick response. I am also considering to employ a robust ML estimator. Thus, when using MLR implement in Mplus, are BIC or AIC of different (nested) models comparable?

Best
Tino
 Bengt O. Muthen posted on Wednesday, September 28, 2016 - 12:57 pm
Yes, BIC/AIC are a function of only the logL and the number of parameters and those are not different for ML and MLR.
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: