Checking assumptions for Ordered Logistic Regression

[Kennedy Nkhoma]

I need help with commands for Brant test of parallel Regression Assumption. Thanks Kennedy

[Maarten Buis]

See ssc desc oparallel

[Oded Mcdossi]

You should also consider (if you haven't already) gologit2 by Richard Williams :
ssc desc gologit2

[Richard Williams]

oparallel is a very good command and gives the brant test and several others. The brant command itself is part of Long and Freese's spost13 package. Type

findit spost13_ado

to find and install it. The main added advantage of the brant command is that it includes a -detail- option that indicates what the most problematic variables are.

gologit2 is potentially useful when the assumptions of the ologit or oprobit command are violated. The support page is at

http://www3.nd.edu/~rwilliam/gologit2/index.html

Also potentially useful with several multiple-outcome commands (e.g. mlogit, ologit, gologit2) is

http://www3.nd.edu/~rwilliam/xsoc73994/Margins05.pdf