Predicted probability and classification statistics for xlogit

[Arbnor Gashi]

Dear Statalist users,

I have two question regarding the fixed effect logit in panel settings.

1) I am trying to get the predicted probablility after using the xtlogit ,fe command and that doesn't work with xtlogit. In time series setting I have been using something like:

logit event var1 var2
predict prob, pr

but when I use the same command after xtlogit event var1 var2, fe doesn't work.

Is there any command to get the predicted probability after xtlogit?


2) Second question has to do with classification statistics and table for xtlogit. Again as in the first case, I used the estat classification after the logit model but this doesn't work for the xtlogit. Does Stata provides any command to report

c

lassification statistics fo xtlogit mode?

Thanks in advance.

[Richard Williams]

See

help xtlogit_postestimation

to see what your options are,

Also, this recent thread may be of interest. Post-estimation after xtlogit, fe, is not straightforward.

https://www.statalist.org/forums/for...tic-regression

[Arbnor Gashi]

Actually, there are two different commands to get the predicted probability after the xtlogit ,fe (I can use either pc1 or pu0). Which of the command is equivalent to pr in a probit model?