The SAS System                                9
                                                 11:39 Wednesday, April 18, 2007

                             The LOGISTIC Procedure

                               Model Information

                 Data Set                      WORK.ICECREAM
                 Response Variable             z2
                 Number of Response Levels     2
                 Weight Variable               Freq
                 Model                         binary logit
                 Optimization Technique        Fisher's scoring


                    Number of Observations Read          72
                    Number of Observations Used          68
                    Sum of Weights Read                 496
                    Sum of Weights Used                 496


                                Response Profile

               Ordered                      Total            Total
                 Value           z2     Frequency           Weight

                     1            1            55        449.00000
                     2            0            13         47.00000

                          Probability modeled is z2=1.

NOTE: 4 observations having nonpositive frequencies or weights were excluded
      since they do not contribute to the analysis.


                            Model Convergence Status

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


                              Model Fit Statistics

                                                  Intercept
                                   Intercept            and
                     Criterion          Only     Covariates

                     AIC             312.903        275.304
                     SC              315.122        281.963
                     -2 Log L        310.903        269.304


                                 The SAS System                               10
                                                 11:39 Wednesday, April 18, 2007

                             The LOGISTIC Procedure

                    Testing Global Null Hypothesis: BETA=0

            Test                 Chi-Square       DF     Pr > ChiSq

            Likelihood Ratio        41.5984        2         <.0001
            Score                   43.9777        2         <.0001
            Wald                    35.4699        2         <.0001


                   Analysis of Maximum Likelihood Estimates

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

      Intercept     1      0.8931      0.2604       11.7621        0.0006
      U             1     39.2798      6.6305       35.0945        <.0001
      U2            1      -132.8     23.4314       32.1438        <.0001


                              Odds Ratio Estimates

                                Point          95% Wald
                   Effect    Estimate      Confidence Limits

                   U         >999.999    >999.999    >999.999
                   U2          <0.001      <0.001      <0.001


         Association of Predicted Probabilities and Observed Responses

               Percent Concordant     49.4    Somers' D    0.112
               Percent Discordant     38.2    Gamma        0.128
               Percent Tied           12.4    Tau-a        0.035
               Pairs                   715    c            0.556