Dear all, I would like to know your opinion about an apparently strange (to me) result:

I run a simple OLS model with an interaction term between a dummy and a continuous variable (see dataex below) on a subset of my sample (only female).
Code:
reg rel_sat aage i.edu i.acountry   c.housetask_ok##i.ftf if male ==0
When I ask the marginal effect with the command
Code:
margins ftf, at(housetask_ok=(0(5)20))
I obtain the following results, that in some cases predict a value higher than 10 (which is the maximum value of my dependent variable):

Code:
------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     _at#ftf |
        1 0  |   10.73738   .2657561    40.40   0.000     10.21576      11.259
        1 1  |   10.03764   .4744374    21.16   0.000     9.106419    10.96886
        2 0  |   9.778896    .155977    62.69   0.000     9.472747    10.08505
        2 1  |   9.571884   .2829458    33.83   0.000     9.016523    10.12725
        3 0  |   8.820414   .0666106   132.42   0.000     8.689671    8.951156
        3 1  |   9.106131     .14082    64.67   0.000     8.829731     9.38253
        4 0  |   7.861931   .1046459    75.13   0.000     7.656534    8.067329
        4 1  |   8.640377   .2107291    41.00   0.000     8.226761    9.053993
        5 0  |   6.903449   .2094521    32.96   0.000     6.492339    7.314558
        5 1  |   8.174624   .3921577    20.85   0.000     7.404902    8.944345
------------------------------------------------------------------------------
How is it possible? Do you have any possible explanation on that?

Thanks a lot in advance,best, G

Code:
* Example generated by -dataex-. To install: ssc install dataex
clear
input byte(acountry aage rel_sat) float(edu male meet_end housetask_ok)
14 39  8 2 1 0 14
14 34 10 2 0 0  .
14 27 10 3 0 0  9
14 49 .b 2 0 0 15
14 41  . 1 1 1  .
14 21  9 2 1 1  .
14 45  . 1 0 0  .
14 37  . 1 0 1  .
14 28  9 3 1 0  .
14 22  . 3 0 0  .
end
label values acountry acountry
label def acountry 14 "Germany", modify
label values aage NoLabel
label values rel_sat TSatisfaction
label def TSatisfaction 10 "Completely satisfied", modify
label def TSatisfaction .b "Refusal", modify