regression for ordinal data?

#1
Hi All,

I am looking at data showing injury to intervertebral discs. Injuries are graded on a 4 point scale with 0 being no damage and 3 being maximum damage. What we seeing is that when there is an injury at one disc, there seems to be an injury in the proximal disc.

What we are trying to do is predict the injury grade to the proximal segment from the that of the initial segment.

If the data was continuous, I would be able to run a regression but this is ordinal data. What is the best approach to do this considering this is ordinal data?

Thank you very much in advance!

Mac
 

hlsmith

Less is more. Stay pure. Stay poor.
#2
Are you familiar with logistic regression? There should be a form of regression (ordinal rgression) that is similar you can use.