In general, most of the students came from magnet schools that you have to test into. These schools sometimes have relationships with local universities or labs, so there are research progams avaliable to the students. I think the average timeframe for the research was about a year, with some people spending more time (the 7th place winner spent 4 years on her research!) and some people spending less. Also, several students worked completely on their own (like the ninth place winner, she built and concieved her project in her basement). IMO those are the most impressive projects. As for the "soft" sciences, in the top forty projects, there were two behavioral and social sciences projects. I think the reason why they didn't win isn't because of the quality of their projects, but the rigourous judging that all the finalists had to go through. They made it clear that they weren't just looking for the best project, but future leaders in science. We had four interview sessions, and the judges also observed us when we were presenting our projects at the NAS to the public.
An anonymous reader writes "I have worked in tech support for the last several years, but find myself wanting to move on to something else — programming. I've written some small programs in my limited spare time but nothing particularly impressive; just functional stuff to make my life easier. I've spent a lot of time recently working through programming books, and feel I'm ready to make the switch in my career. That said, I don't have a CS degree, and find that responses to my resume have been along the lines of "Thanks, but we aren't hiring for tech support positions." Surely someone from the slashdot crowd has been in the same position — what would you recommend?"