Ordinal probit regression and dummy variables

Moderators: statman, Analyst Techy, andris, Fierce, GerineL, Smash

Posts: 1
Joined: Wed Feb 24, 2016 2:56 pm

Ordinal probit regression and dummy variables

Postby Cocogiraf » Wed Feb 24, 2016 3:09 pm

Hi, I am trying to run an ordinal probit regression however I am struggling with several things:

1) Does the dependant variable have to be a dummy variable? Currently it has three options = 0 -No, 1- yes in the past, and 2- yes currently. Therefore can I run it like that or does it have to be 0 and 1 only?

2) Do my independent variables (that are categorical) have to be dummy variables? (similar case as for above).

3) If they have to be dummy variables - when I create dummy variables in spss, there are two variables for each case then, do I include them both in the regression?

Also, I have read that because I have several variables that are numerous (scale), I should run Ordinal regression and choose Probit in Options. Does anyone know if that's correct?

Any help would be appreciated as I am feeling lost. Thank you.

P.S. I use spss v22 if that's of any help.
Posts: 1477
Joined: Tue Jun 10, 2008 4:50 pm

Re: Ordinal probit regression and dummy variables

Postby GerineL » Thu Feb 25, 2016 12:59 pm

Ordinal probit regression, or ordered logistic regression, is usually meant to examine relations between one or more ordinal variables, and a binary outcome variable.
If you have ordinal dependent and ordinal independent variables, you could for instance look at multinomial regression.
I don't understand your third question.

Who is online

Users browsing this forum: No registered users and 1 guest