SHAZAM Community
Ask Your Question
0

How to interpret ordered logistic regression

asked 2014-02-18 09:00:30 +0000

Madelaine gravatar image

Dear all,

My name is Madelaine, I am doing research on party list rankings for open list PR elections (where votes are cast on a single candidate). Using a range of explanatory variables (age, education, effort in parliament etc.) I try to estimate the rank a candidate has on a party list.

As my dependent variabele (rank) is ordinal, I wanted to use an ordered logistic panel data regression. However, since the rank can take up values ranging from 1 to 80, I do not know how to interpret the odd ratios that come out of the regression. I also doubt whether the ordered logistic regression model is the correct model to use when the dependent rank variable has 80 categories.

I would very much appreciate it if someone could help me out on this,

Many thanks in advance, and kind regards,

Madelaine.

edit retag flag offensive close merge delete

Comments

Hi, if you just want to estimate how a set of characteristics (explanatory variables) affect the ranking, why not run an OLS-regression?

Pantera gravatar imagePantera ( 2014-02-27 16:08:52 +0000 )edit

1 answer

Sort by ยป oldest newest most voted
0

answered 2014-02-20 02:49:16 +0000

DrBob gravatar image

I suggest you look at the ordered logistic SHAZAM routines prepared by Daehoon Nahm at Macquarie SHAZAM Ordered and Multinomial Logit Procedures. His routine prints out the coefficients (and the prediction success table) and you can calculate the odds ratio afterwards from the output by exponentiating the ordered logit coefficients.

Ordered logit is estimation over the levels of your dependent variable (rank) so the odds ratio is comparing the votes of people who are in groups greater than k versus those who are in groups less than or equal to k, with k being the level of the response variable. The interpretation would be that for a one unit change in the particular predictor variable (e.g. age), the odds for cases in a group that is greater than k versus a group that is less than or equal to k are the (proportional odds) times larger. e.g. if the Odds ratio of a variable e.g. age are say 3.2 then the odds of the the vote being for the the rank > K would be just over 3 times larger.

edit flag offensive delete link more
Login/Signup to Answer

Question Tools

Stats

Asked: 2014-02-18 09:00:30 +0000

Seen: 1,072 times

Last updated: Feb 20 '14