Negative Chi Square PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
Message/Author
 Eeske van Roekel posted on Friday, February 08, 2013 - 12:19 am
Hello,
Because I use MLR in my models, I estimate the difference in model fit between subsequent models with the Satorra Bentler chi-square test. However, when comparing fits between two models, the chi square turned out negative. Therefore I tried calculating the strictly positive chi square test, as mentioned on your website. I have created the M10 model, but there are some problems with the iterations. I have tried both the convergence=100000000 and the MITER=1 options (separately and simultaneously), but it seems Mplus is still taking some iterations, because the slopes (which should be 0) are still estimated and the loglikelihood of the model is the same as that for the M1 model (when I use both options, the slopes are 0, but the loglikelihood is only slightly higher than that of the M1 model).
Do you know what the problem is?
Thanks in advance,
Eeske van Roekel
 Linda K. Muthen posted on Friday, February 08, 2013 - 11:49 am
Please send the relevant files and your 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: