The intercepts of a cumulative logit model

[bncoxuk]

Hi,

I used PROC LOGISTIC to run a cumulative logit model for ordinal response. The dependent variable has 3 levels:  group=1, group=2, and group=3. For the cumulative logit model here, the predicition is for the probability in the lower category (group 1 vs group 2, 3,  and group 1, 2 vs group 3).

I don't undertand the meaning of the intercepts, which are produced as:

_LABEL_

Intercept: group=3     2.325

Intercept: group=2     5.273

I am very confused with the meaning of intercepts here. Can you please help?

[Reeza]

Have you tried looking at the Cheese example in the SAS Doc? Example 51.3 in SAS 9.2 documentation, similar situation or are you unsure how to specifically interpret your results?