Hello everyone,
I am currently analyzing my data by running a Multinomial Logistic Regression with 7 response categories without natural ordering (dependent variable) and a binary coded (independent) variable. Moreover I have some controls that are all continuous (like ROA or firm age). I found a paper that has a similar method that states they are using a "seemingly unrelated estimation-based Hausman" to check the assumption that "that the relative risk ratio between any pair of outcome categories does not depend on the other alternatives available". I am looking for a way how to perform this test in stata.
Thanks a lot,
Kim
I am currently analyzing my data by running a Multinomial Logistic Regression with 7 response categories without natural ordering (dependent variable) and a binary coded (independent) variable. Moreover I have some controls that are all continuous (like ROA or firm age). I found a paper that has a similar method that states they are using a "seemingly unrelated estimation-based Hausman" to check the assumption that "that the relative risk ratio between any pair of outcome categories does not depend on the other alternatives available". I am looking for a way how to perform this test in stata.
Thanks a lot,
Kim
