st: how to graph predicted probabilities with prgen

[Ekaterina Hertog <ekaterina.hertog@sociology.ox.ac.uk>]

Dear all,
I use Stata 12 and I am analysing the effect of income on overtime work 
using the following model:

logit above40hrs age agesq uni income incsq

I want to see how the predicted probabilities of working more than 40 
hours a week depend on individual income using prgen:

I tried the following:
prgen income, from(100) to(1100) generate(prab40jf12) x(age=30
agesq=900 uni=1)  ci

I got the following output:

logit: Predicted values as income varies from 100 to 1100.

          age      agesq        uni    income     incsq
x=         30        900          1  380.12523   184506.9

It is immediately obvious that the value displayed for incsq (income 
squared) is much more than 380*380, so I am worried that when Stata 
estimates its predictions it uses incorrect values for incsq (Indeed 
when I use prvalues to make individual predictions and set the values of 
income and incsq manually I get somewhat different predictions). How 
would I use prgen in such a case?
I know that it is more typical to use logged income, but I am trying to 
test a hypothesis that low income and high income men work particularly 
long hours with everyone else working less.

I will be very grateful for advice,
Sincerely yours,
Katya
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/