How to remove outliers from logistic regression?

[MikeTurner]

Outliers in my logistic model suffered me a lot these days. Anyone has some experience at this?

And, by the rule of thumb, what value of hit rate could be considered a satisfactory result (I have four nominal dependent variables in my model)? Are there some reference papers?

Any help will be appreciated. Thanks!

[Rick_SAS]

Look at this post for ways to identify outliers: https://communities.sas.com/message/113376#113376