Predicted Probablities for ZINB PreviousNext
Mplus Discussion > Categorical Data Modeling >
 Rich Mohn posted on Tuesday, January 27, 2015 - 10:19 am

I am running my first zero inflated negative binomial regression, and of course, having fun : ). I would like to get predicted probabilities for the logistic part of the model to see who is likely to go from a '0' to a value of 1 or more. I didn't see a way to save them, but can I just calculate it just like I'd do for a logistic regression? Thanks for any advice.
 Bengt O. Muthen posted on Tuesday, January 27, 2015 - 3:26 pm
I think so since no random effect enters into this part of the model. I assume you are talking about the inflation equation (being able to only observe zero or not).
 Rich Mohn posted on Wednesday, January 28, 2015 - 8:59 am
Yep, thanks.
 Rich Mohn posted on Thursday, February 05, 2015 - 9:32 am
Hello again . . . using only the logistic estimates doesn't seem to be working, and I'm guessing it's because I'm not using information from the other model.

Is there a way to save predicted probabilities in a file? I've looked through the manual and tried a few things, but have not been successful.

 Bengt O. Muthen posted on Friday, February 06, 2015 - 9:50 am
Note that the logistic part predicts the inflation (the class of people who can only have zero), so not the probability of observing a zero. The non-zero class people also have a probability of zero.

I don't think we save those probabilities.
 Bengt O. Muthen posted on Friday, February 06, 2015 - 10:22 am
Actually, we provide plots for both the count and the binary (inflated) part if you have the inflated part mentioned in the SERIES option. Which means that you can save those values.
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