|
|
Non positive definite and fit statistics |
|
Message/Author |
|
Peter Ji posted on Wednesday, February 10, 2010 - 11:17 am
|
|
|
I ran some models and for some of the them I received a NPD, non positive definite. So the model's results cannot be interpreted. But I even if the model was NPD, the output still presents fit statistics. Why is that so? and even when MPLUS produces fit statistics with a NPD, the model is still uninterpretable, correct? |
|
|
A model with inadmissible parameter values, such as NPD, can still fit the data well. But it is not defensible due to the NPD. |
|
Back to top |
|
|