Dear all, thank you in advance for your time.
I conducted an ordinal regression using gologit2 with an outcome that has 4 response levels.
Not knowing how to perform a good diagnostic, I thought of using estat gof between one level and another.
Should I use estat gof between level 1 and level 2, then between level 2 and level 3, and so on?
OR
Should I compare level 1 with levels 2, 3, and 4, then compare levels 1 and 2 with levels 3 and 4, and finally compare levels 1, 2, and 3 with level 4?
Please note that gologit2 relaxes the assumption of proportionality (it is a semiproportional model). This is why I have progressive aggregated estimates.
Thanks again.
Gianfranco
I conducted an ordinal regression using gologit2 with an outcome that has 4 response levels.
Not knowing how to perform a good diagnostic, I thought of using estat gof between one level and another.
Should I use estat gof between level 1 and level 2, then between level 2 and level 3, and so on?
OR
Should I compare level 1 with levels 2, 3, and 4, then compare levels 1 and 2 with levels 3 and 4, and finally compare levels 1, 2, and 3 with level 4?
Please note that gologit2 relaxes the assumption of proportionality (it is a semiproportional model). This is why I have progressive aggregated estimates.
Thanks again.
Gianfranco
