The SAS System          08:54 Wednesday, August 8, 2001  25

                                          The Mixed Procedure

                                           Model Information

                         Data Set                     WORK.ANTI
                         Dependent Variable           anti
                         Covariance Structure         Unstructured
                         Subject Effect               subjid
                         Estimation Method            REML
                         Residual Variance Method     Profile
                         Fixed Effects SE Method      Model-Based
                         Degrees of Freedom Method    Between-Within


                                              Dimensions

                                  Covariance Parameters             4
                                  Columns in X                      6
                                  Columns in Z Per Subject          2
                                  Subjects                        221
                                  Max Obs Per Subject               4
                                  Observations Used               884
                                  Observations Not Used             0
                                  Total Observations              884


                                           Iteration History

                      Iteration    Evaluations    -2 Res Log Like       Criterion

                              0              1      3516.52765660
                              1              1      3281.65500131      0.00000000


                                       Convergence criteria met.


                                     Estimated G Correlation Matrix

                           Row    Effect       subjid        Col1        Col2

                             1    Intercept      1         1.0000      0.4209
                             2    time           1         0.4209      1.0000


                                     Covariance Parameter Estimates

                                                       Standard         Z
                   Cov Parm     Subject    Estimate       Error     Value        Pr Z

                   UN(1,1)      subjid       0.8851      0.2007      4.41      <.0001
                   UN(2,1)      subjid       0.1173     0.07083      1.66      0.0977
                   UN(2,2)      subjid      0.08774     0.04294      2.04      0.0205
                   Residual                  1.5290      0.1028     14.87      <.0001
                                             The SAS System          08:54 Wednesday, August 8, 2001  26

                                          The Mixed Procedure

                                            Fit Statistics

                                 -2 Res Log Likelihood          3281.7
                                 AIC (smaller is better)        3289.7
                                 AICC (smaller is better)       3289.7
                                 BIC (smaller is better)        3303.2


                                    Null Model Likelihood Ratio Test

                                      DF    Chi-Square      Pr > ChiSq

                                       3        234.87          <.0001


                                      Solution for Fixed Effects

                                              Standard
                  Effect          Estimate       Error      DF    t Value    Pr > |t|

                  Intercept         1.8240      0.3757     218       4.86      <.0001
                  time              0.5465      0.1685     660       3.24      0.0012
                  gen               0.5716      0.1884     218       3.03      0.0027
                  homecog         -0.06261     0.03840     218      -1.63      0.1045
                  time*gen         0.07910     0.08450     660       0.94      0.3496
                  time*homecog    -0.04523     0.01723     660      -2.62      0.0089