Baseline logit model                       1

                        The CATMOD Procedure

                        Population Profiles
 
                  Sample    contact    Sample Size
                  --------------------------------
                      1     High               968
                      2     Low                713

                         Response Profiles
                        Response    rsat
                        --------------------
                            1       High    
                            2       Intermed
                            3       Low     

              Analysis of Maximum Likelihood Estimates
 
             Function          Standard   Chi-
 Parameter    Number   Estimate   Error Square  Pr > ChiSq
 ---------------------------------------------------------
 Intercept      1        0.1499  0.0576   6.76      0.0093
                2       -0.2579  0.0641  16.18      <.0001
 contact   High 1        0.1087  0.0576   3.56      0.0593
           High 2        0.1286  0.0641   4.02      0.0449
               Binary response for comparison purpose 

                       The LOGISTIC Procedure

                         Model Information

                  Probability modeled is oksat=0.
                      Class Level Information
                                          Design
                  Class       Value     Variables

                  contact     High              1
                              Low              -1

             Analysis of Maximum Likelihood Estimates
 
                             Standard         Wald
Parameter     DF  Estimate      Error   Chi-Square  Pr > ChiSq

Intercept      1   -0.6598     0.0520     160.9178      <.0001
contact   High 1   -0.1167     0.0520       5.0317      0.0249


                    Odds Ratio Estimates
                              
                             Point          95% Wald
   Effect                 Estimate      Confidence Limits
   contact High vs Low       0.792       0.646       0.971

               Binary response for comparison purpose 
                        The GENMOD Procedure
                         Model Information

                      Class Level Information
 
                  Class        Levels    Values

                  contact           2    High Low 


                          Response Profile
 
                    Ordered                 Total
                      Value    oksat    Frequency

                          1    0              567
                          2    1             1114

PROC GENMOD is modeling the probability that oksat='0'. One
way to change this to model the probability that oksat='1' 
is to specify the DESCENDING option in the PROC statement.

        Analysis Of Maximum Likelihood Parameter Estimates
                                 Standard Wald 95% Confidence
 Parameter        DF   Estimate   Error        Limits
 Intercept         1    -0.5431  0.0777  -0.6954    -0.3909
 contact   High    1    -0.2333  0.1040  -0.4372    -0.0295
 contact   Low     0     0.0000  0.0000   0.0000     0.0000
 Scale             0     1.0000  0.0000   1.0000     1.0000

                  Analysis Of Maximum Likelihood
                        Parameter Estimates
 
                                     Wald
            Parameter          Chi-Square    Pr > ChiSq

            Intercept               48.89        <.0001
            contact     High         5.03        0.0249
            contact     Low           .           .    
            Scale                                      

               Binary response for comparison purpose

                         The FREQ Procedure

                                   Cumulative    Cumulative
   rsat   Frequency     Percent     Frequency      Percent
   --------------------------------------------------------
   High        668       39.74           668        39.74  
   Intermed    446       26.53          1114        66.27  
   Low         567       33.73          1681       100.00  
            Ordinal Logistic Regression with wrong order

                       The LOGISTIC Procedure

                         Model Information

           Data Set                      WORK.HOUSING    
           Response Variable             sat             
           Number of Response Levels     3               
           Frequency Variable            freq            
           Model                         cumulative logit
           Optimization Technique        Fisher's scoring


              Number of Observations Read          72
              Number of Observations Used          72
              Sum of Frequencies Read            1681
              Sum of Frequencies Used            1681


                          Response Profile
 
                 Ordered                      Total
                   Value     sat          Frequency

                       1     High               668
                       2     Low                567
                       3     Medium             446

 Probabilities are cumulated over the lower Ordered Values.


                      Model Convergence Status

           Convergence criterion (GCONV=1E-8) satisfied.          

                        -2 Log L = 3648.878 


             Analysis of Maximum Likelihood Estimates
 
                              Standard      Wald
Parameter      DF   Estimate   Error   Chi-Square Pr > ChiSq
Intercept High  1    -0.4164  0.0498      69.7919     <.0001
Intercept Low   1     1.0185  0.0552     339.9088     <.0001
        Ordinal Logistic Reg. with right order, empty model

                       The LOGISTIC Procedure

 Probabilities are cumulated over the lower Ordered Values.

                                 Standard        Wald Pr >
 Parameter         DF  Estimate   Error  Chi-Square   ChiSq

 Intercept Low      1   -0.6754  0.0516    171.3809    <.0001
 Intercept Inter    1    0.4164  0.0498     69.7919    <.0001
             Ordinal Logist. Reg. with added covariates

                       The LOGISTIC Procedure

                      Class Level Information
 
             Class       Value        Design Variables

             infl        High          1      0       
                         Low           0      1       
                         Medium       -1     -1       

             type        Apartmen      1      0      0
                         Atrium        0      1      0
                         Terrace       0      0      1
                         Tower        -1     -1     -1

             contact     High          1              
                         Low          -1              

           Score Test for the Proportional Odds Assumption
 
                 Chi-Square       DF     Pr > ChiSq
                   103.3500        6         <.0001

              Testing Global Null Hypothesis: BETA=0
 
      Test                 Chi-Square       DF     Pr > ChiSq
      Likelihood Ratio        75.7241        6         <.0001
      Score                   73.2308        6         <.0001
      Wald                    73.7700        6         <.0001


                    Type 3 Analysis of Effects
 
                                     Wald
            Effect       DF    Chi-Square    Pr > ChiSq

            infl          2       54.0653        <.0001
            type          3       19.5243        0.0002
            contact       1        2.4578        0.1169

              Analysis of Maximum Likelihood Estimates
                               Standard        Wald  Pr >
 Parameter         DF  Estimate   Error  Chi-Square  ChiSq

 Intercept Medium   1   -1.1085  0.0624    315.7672   <.0001
 Intercept Low      1    0.3777  0.0567     44.3976   <.0001
 infl      High     1   -0.5261  0.0744     49.9529   <.0001
 infl      Low      1    0.3747  0.0642     34.0360   <.0001
 type      Apartmen 1   -0.0350  0.0693      0.2547   0.6138
 type      Atrium   1    0.1282  0.0988      1.6823   0.1946
 type      Terrace  1    0.2589  0.0936      7.6510   0.0057
 contact   High     1   -0.0734  0.0469      2.4578   0.1169


                     Odds Ratio Estimates
                                 Point          95% Wald
 Effect                       Estimate      Confidence Limits
 infl    High vs Medium          0.508       0.400       0.645
 infl    Low  vs Medium          1.250       1.021       1.531
 type    Apartmen vs Tower       1.373       1.092       1.727
 type    Atrium   vs Tower       1.617       1.195       2.188
 type    Terrace  vs Tower       1.842       1.380       2.460
 contact High vs Low             0.863       0.719       1.037

             Complimentary log log                     

                      Class Level Information
 
             Class       Value        Design Variables

             infl        High          1      0       
                         Low           0      1       
                         Medium       -1     -1       

             type        Apartmen      1      0      0
                         Atrium        0      1      0
                         Terrace       0      0      1
                         Tower        -1     -1     -1

             contact     High          1              
                         Low          -1              

             Score Test for the Equal Slopes Assumption
                 Chi-Square       DF     Pr > ChiSq
                    73.9313        6         <.0001

              Testing Global Null Hypothesis: BETA=0
 
      Test                 Chi-Square       DF     Pr > ChiSq

      Likelihood Ratio       107.3603        6         <.0001
      Score                  106.0977        6         <.0001
      Wald                    99.9839        6         <.0001

                    Type 3 Analysis of Effects
                                     Wald
            Effect       DF    Chi-Square    Pr > ChiSq

            infl          2       68.2525        <.0001
            type          3       29.6460        <.0001
            contact       1        4.3575        0.0368


              Analysis of Maximum Likelihood Estimates
 
                                Standard Wald
Parameter         DF  Estimate  Error  Chi-Square  Pr > ChiSq

Intercept Medium   1   -1.2570  0.0525    573.3420     <.0001
Intercept Low      1   -0.1235  0.0386     10.2263     0.0014
infl      High     1   -0.4251  0.0568     56.0243     <.0001
infl      Low      1    0.3334  0.0451     54.7552     <.0001
type      Apartmen 1   0.00372  0.0484      0.0059     0.9387
type      Atrium   1    0.0275  0.0690      0.1587     0.6903
type      Terrace  1    0.2619  0.0625     17.5789     <.0001
contact   High     1   -0.0685  0.0328      4.3575     0.0368