I have too many women in my sample - stoopid women (or men for not doing my questionnaire).

If I control for gender in my regression, can I then generalise results to the population? or are my results just rubbish because the sample is not representative??

(sprout)