Hello,
I am trying to test the independence of irrelevant alternatives (IIA) assumption required for the multinomial logit specification.
I know that this is generally tested using a standard Hausman test, but it seems highly likely that my panel data is characterized by clustered standard errors, which should make the normal Hausman format inappropriate. Cameron and Trivedi have some discussion of how to do a robust variation of the Hausman (Section 21.4.3), but they consider the issue in the context of random effects vs POLS vs FE etc. Moreover, the statistical example they give just has one regressor, which is not very helpful for me.
Any thoughts or suggestions would be very much appreciated.
Thanks,
Nathan Wilson
PhD Student
Business Economics
Ross School of Business
University of Michigan
�
[email protected]
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/