On Jennifer Crusie's Argh Ink blog, she had a recent post, "The Wanderer's Guild To Story Evolution" , about the thought processes involved in plotting out a story. Great stuff, as is often the case with Crusie. She goes into very amusing depth about a potential project and the various choices and thoughts she's had so far.
I resemble a lot of the things she mentions, in my own writing, especially when I'm trying to write at longer length. Usually start with an image: Who are these people? Why are they there? Why are they doing what they're doing? All very fuzzy and loose at first, a mystery even to the author.
Build from there. Trying out different possibilities and ideas. Keeping some, discarding others, keeping others in on a provisional basis "for a reason to be named later", or not. That initial fuzzy view gets a little clearer, a little more focused, a little more sensible each time you go back in to give it more thought. Do the new ideas work? Are the characters consistent? Do their new actions make sense with what's already written? If not, revise the old writing or discard the new?
And particularly, that sometimes you have to walk away for a bit, let the story perk away in your subconscious for a while. In Crusie's parlance, let the dough rise before you go back to punch it down and work it some more.
This last is why I'll probably never participate in Nanowrimo. For me, I seem to get better (if slower) results by not forcing myself to write X number of words per day, or to plot according to a strict structure, or by making myself finish one story before starting another.
(It's easier to get away with this when you're only a part-time writer. For Crusie, a professional, it means keeping multiple projects in the air, some at the point of contracts and advances, some not.)
I've been working on... I hesitate to use the N-word, because I've never finished a novel, but... "a longer work", and progress on that has been more a process of punctuated equilibrium than steady evolution. I've taken breaks from that longer work, and written several short stories instead, in recent months. And I think that's been useful. I started out with a fairly strong idea of the backstory and several main characters, what kind of story I wanted it to be, a fairly solid opening image, a rough idea of what the ending would be, and a *B*I*G* *F*U*Z*Z*Y* of everything in between. Those breaks let the story perk in my backbrain, and I think the results have been better than if I'd tried to force the wordcount up in a quicker manner. That *B*I*G* *F*U*Z*Z*Y* is now more of a *M*E*D*I*U*M*-*S*I*Z*E* *F*U*Z*Z*Y* and the choices -- choices that feel like the right choices -- of what needs to happen in the story are coming more quickly and easily.
(It's kind of like working on a jigsaw puzzle after only hearing a brief description of the box illustration. Start with the easier edge pieces. Study the confusing jumble of interior pieces. Find colors and patterns that are similar. Group them together. Twist and turn the pieces, seeing which match and which don't. The more pieces that are fitted in, the fewer left, and the easier it becomes to match and fit those until the puzzle is finally complete.)
Check Crusie's piece out; it's one of the best descriptions of "Writer's Mind" I've seen.