Richard Williams' very neat -gologit2- should meet your needs. It
supports iweights, fweights and pweights, and it goes beyond just
testing for the parallel regression assumption by allowing you to fit
models which violate it. There's also a very good discussion paper:
http://www.nd.edu/~rwilliam/stata/gologit2.pdf
The program is available from ssc archives by typing:
ssc install gologit2
Thanks to Ian for the plug, especially since I had failed to read the
original message. The program also supports the Stata 8.2 svy
features if your data set is set up that way (the new features in 9.0
are not supported yet.) More information about the program, as well
as a troubleshooting FAQ, can be found at
