Proportional odds regression model for global impression scale at month 6
Arthritis Clinical Trial

The GENMOD Procedure

Model Information
Data Set WORK.ARTHRITIS
Distribution Multinomial
Link Function Cumulative Logit
Dependent Variable y4

Number of Observations Read 303
Number of Observations Used 293
Missing Values 10

Response Profile
Ordered
Value		 y4 Total
Frequency
1 1 38
2 2 93
3 3 103
4 4 49
5 5 10


PROC GENMOD is modeling the probabilities of levels of y4 having LOWER Ordered Values in the response profile table. One way to change this to model the probabilities of HIGHER Ordered Values is to specify the DESCENDING option in the PROC statement.

Criteria For Assessing Goodness Of Fit
Criterion DF Value Value/DF
Log Likelihood   -407.0547  
Full Log Likelihood   -407.0547  
AIC (smaller is better)   826.1095  
AICC (smaller is better)   826.4032  
BIC (smaller is better)   848.1905  

Algorithm converged.

Analysis Of Maximum Likelihood Parameter Estimates
Parameter DF Estimate Standard Error Wald 95% Confidence Limits Wald Chi-Square Pr > ChiSq
Intercept1 1 -1.2118 0.5316 -2.2537 -0.1698 5.20 0.0226
Intercept2 1 0.5251 0.5249 -0.5036 1.5539 1.00 0.3171
Intercept3 1 2.1494 0.5381 1.0948 3.2040 15.96 <.0001
Intercept4 1 4.1366 0.6121 2.9368 5.3363 45.67 <.0001
age 1 -0.2048 0.0983 -0.3975 -0.0121 4.34 0.0372
trt 1 0.6079 0.2142 0.1880 1.0278 8.05 0.0045
Scale 0 1.0000 0.0000 1.0000 1.0000    

Note: The scale parameter was held fixed.