NPD warning PreviousNext
Mplus Discussion > Structural Equation Modeling >
Message/Author
 Abby Gray posted on Wednesday, January 08, 2014 - 6:53 am
Are there any circumstances under which Not Positive Definite warning can or should be ignored, or is this always indicative of serious problems with the model?

Thank you.
 Abby Gray posted on Wednesday, January 08, 2014 - 7:50 am
To my message above, I meant to add that the non-standardized results show correlations greater than one between some indicators for one of my latent variables. I assume this is triggering the warning. However, I plan to use only standardized output, and in the STDYX results there are no correlations above zero. In this case, is it OK to ignore the warning?
 Linda K. Muthen posted on Wednesday, January 08, 2014 - 9:50 am
This warning should not be ignored. Correlations greater than one make the model inadmissible. The standardized results are also inadmissible. You need to change your model.
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: