Inverse prediction for logistic and probit regression

[Ray Webster <WebsterR@landcare.cri.nz>]

Has anyone written functions for inverse prediction for logitistic or probit
regression models?  I'm fitting dose-response models, where the
response is whether the subject dies or not. (The subject is an exotic
pest species, by the way.)  We'd like to be able to obtain estimated LCXs
- the concentration at which X% of the population dies - and confidence
intervals for these.

Thanks

Ray Webster
Landcare Research
Lincoln
New Zealand