|
|
 |
| Overdispersion with logistic regression |
 
|
| Message/Author |
|
|
| Anonymous posted on Wednesday, March 27, 2013 - 8:34 am
|
|
|
| I have a number of models with categorical outcomes (logistic, multinomial and ordinal) which I'd like to test for overdisperson. Is this possible in Mplus? |
|
|
|
|
There isn't an automatic way of doing this in those cases, but you can probably use TYPE=MIXTURE to do it by adding a latent class of people who can only have a zero value.
UG ex 7.25 does zero-inflated Poisson using such an approach. |
|
|
| Back to top |
|
|
