Previous Chapter Chapter 75: Proportional Odds Model Next Chapter

  • Citations
  • Add to My List
  • Text Size

Proportional Odds Model
Proportional odds model

The generalized linear modelling technique of proportional odds allows ordered categorical response variables to be modelled using a logit regression analysis. Ordered categorical variables are common in management research and can indicate, for example, levels of agreement (strongly agree to strongly disagree) to seniority in decision making and educational achievement. The proportional odds model is a logit model that can be viewed quite simply as an extension of logistic regression. This technique allows ordered data to be modelled by analysing it as a number of dichotomies. A binary logistic regression model compares one dichotomy (for example, passed–failed, died– survived, etc.), whereas the proportional odds model compares a number of dichotomies by arranging the ordered categories into a series of binary comparisons. ...

Looks like you do not have access to this content.


Don’t know how to login?

Click here for free trial login.

Back to Top

Copy and paste the following HTML into your website