We intend to do this by operating a program within our company, starting with community computer and science lab, promoting to a group work day with interactive collaboration, then eventually fitting a child's interest in a role with a second-seat internship at each employee's desk.
This is all early stage of process, but 20 years of thinking and experimentation. If I'm able to build my current venture out, my next will be promoting these early programs to help expand young people's thinking (and hopefully help them make themselves their own role model) about science and technology.
http://www.prb.org/Articles/2007/CrossoverinFemaleMaleColleg...
http://nces.ed.gov/programs/coe/2009/section1/indicator11.as...
To wit: Females have dominated academia for a couple decades now.
Also, it should be pointed out that many founders are college dropouts or even high school dropouts, so I'd be interested in some sort of citation of evidence that one can assume most founders will be a subset of academia.
I know in my college, in my two majors (mechanical engineer & computer science, 2007-2010) it was about 80/20 male/female in ME and 70/30 male/female in CS. The girls average was likely higher than the guys. I don't see how they are "dominating". You're numbers show there are more guys than girls in college - which was also true of my college which overall was 60/40 female/male. That doesn't mean there are more girls in majors related to startups like the ones we are talking about here (which I would highly doubt you can find numbers to say that).
As far as investors go, I doubt that they have any aversion to supporting entrepreneurs of any background. This is one industry that I would presume is more of a meritocracy than society at large. As I said on my tumblr a while back (http://bit.ly/g7jCAF) investors can see the green inside all of us :-)
There are many discussions going on about women in technology startups. The greater question is not why there aren't more founders, but why there arent more minority/female students in academia. Most of your founders are going to be a subset of those in academia anyway, so why not look at the problem closer to the source?