Age Groups, interaction with sex category PLUS Education dummies, i.e.:
 MODEL: 
CON2= A_M_A1*WIN2*(SEX=M)*(AGE_GROUP=1)+ ... +A_M_A6*WIN2*(SEX=M)*(AGE_GROUP=6)+
    A_F_A1*WIN2*(SEX=F)*(AGE_GROUP=1) + ... + A_F_A6*WIN2*(SEX=F)*(AGE_GROUP=6)+
    A_Q1*(EDUC=1)+...+A_Q4*(EDUC=4)

 Number of observations = 1039  E'PZ*E = .165230E-31

                          Standard
 Parameter  Estimate        Error       t-statistic   P-value
 A_M_A1     .276874       .134868       2.05292       [.040]
 A_M_A2     .490741       .104111       4.71364       [.000]
 A_M_A3     .160503       .031157       5.15147       [.000]
 A_M_A4     .301449       .138742       2.17273       [.030]
 A_M_A5     .307121       .097572       3.14765       [.002]
 A_M_A6     .560480       .159822       3.50690       [.000]
 A_F_A1     .341129       .144644       2.35840       [.018]
 A_F_A2     .441724       .126228       3.49941       [.000]
 A_F_A3     .261268       .120192       2.17376       [.030]
 A_F_A4     .355815       .120095       2.96277       [.003]
 A_F_A5     .365515       .133332       2.74139       [.006]
 A_F_A6     .129607       .074225       1.74614       [.081]
 A_E1       .791710       1.17987       .671016       [.502]
 A_E2       -.334958      1.03623       -.323246      [.747]
 A_E3       -1.63816      1.22717       -1.33491      [.182]
 A_E4       -2.63694      1.62238       -1.62535      [.104]

 Standard Errors computed from   heteroscedastic-consistent matrix 
 (Robust-White)

 Dependent variable: CON2
        Mean of dep. var. = 4.60932  Std. error of regression = 8.25252
   Std. dev. of dep. var. = 9.07813                 R-squared = .185562
 Sum of squared residuals = 69670.5        Adjusted R-squared = .173620
    Variance of residuals = 68.1041             Durbin-Watson = 1.96632 [<.550]