Logistic Regression Using SAS®: Theory and Application

6.4. Cumulative Logit Model: Practical Considerations

Now we’re ready to interpret the coefficients in Output 6.1. Because I didn’t use the DESCENDING option, the model predicts the probability of being in a lower category and, in this case, that means less ethical behavior. Each reported odds ratio can be interpreted as the effect of the variable on the odds of being in a lower rather than in a higher category, without regard to how you dichotomize the outcome. For example, the adjusted odds ratio for males is 2.897. We can say, then, that the odds of males being in a lower category (rather than a higher category) is nearly three times the odds for females. Among those in the business school, the odds of being in a lower category are a little ...