Hello!
I use the brant command to test the parallel-line assumption after runing the ordered logistic regression. As the result shows, the parallel-line assumption is violated. So i run the gologit2 command with autofit command. However, the result of gologit2 show that all explanatory variables meet the parallel-line assumption.
Should i directly use the ordered logistic regression or use the gologot2 with npl command?
Thank you very much!
I use the brant command to test the parallel-line assumption after runing the ordered logistic regression. As the result shows, the parallel-line assumption is violated. So i run the gologit2 command with autofit command. However, the result of gologit2 show that all explanatory variables meet the parallel-line assumption.
Should i directly use the ordered logistic regression or use the gologot2 with npl command?
Thank you very much!