Iteractions sex and education categories, i.e.,

 MODEL: CON1= A_M_E1*WIN1*(SEX=M)*(EDUC=1) + ... + A_M_E4*WIN1*(SEX=M)*(EDUC=4)+
              A_F_A1*WIN1*(SEX=F)*(EDUC=2) + ... + A_F_A6*WIN1*(SEX=F)*(EDUC=4)


                                      TOBIT ESTIMATION
      Dependent variable: CON1

  Number of observations = 456  Schwarz B.I.C. = 1460.37
 Number of positive obs. = 331  Log likelihood = -1435.88
 Fraction of positive obs. =    0.725877    

                           Standard
 Parameter   Estimate        Error       t-statistic   P-value
 WIN1SEXME1  .674239       .192734       3.49828       [.000]
 WIN1SEXME2  .543423       .071632       7.58630       [.000]
 WIN1SEXME3  .172938       .029137       5.93525       [.000]
 WIN1SEXME4  .393988       .073554       5.35641       [.000]
 WIN1SEXFE1  .514718       .106698       4.82406       [.000]
 WIN1SEXFE2  .281776       .072825       3.86923       [.000]
 WIN1SEXFE3  .503026       .072269       6.96046       [.000]
 WIN1SEXFE4  .715841       .147372       4.85737       [.000]
 SIGMA       13.0845       .536675       24.3807       [.000]

 Standard Errors computed from   analytic second derivatives 
 (Newton)

                                      dP/dX

                        0             1 
 WIN1SEXME1     -0.017478      0.017478 
 WIN1SEXME2     -0.014087      0.014087 
 WIN1SEXME3    -0.0044830     0.0044830 
 WIN1SEXME4     -0.010213      0.010213 
 WIN1SEXFE1     -0.013343      0.013343 
 WIN1SEXFE2    -0.0073044     0.0073044 
 WIN1SEXFE3     -0.013040      0.013040 
 WIN1SEXFE4     -0.018556      0.018556