One of the really odd things that you may find yourself doing in grad school, is writing a paper before you’ve even run an experiment. Why? Because, if you mainly publish in conferences (which is common in Computer Science), the deadlines happen once a year, and often seem to sneak up on you. Add to that, there’s always a limited number of conferences/journals where your topic fits, and so each deadline is important if you’re going to get stuff published.
For me, the conference best suited for my work to be published in, has a deadline in just over a month. And, as I’ve stated in other posts, I’m still busy trying to sort out the last few details of my experiment (I swear I’m getting closer) and so I haven’t run it yet. Which means I don’t have any results yet.
Sure, I could wait until I have the experiment done, and then write a paper. This may come together before the deadline, or it may not. But, by working on the paper in conjunction with getting the experiment ready, I’m a) giving myself enough time to write and edit a good paper, and b) helping to clarify my thoughts and focus for my experiment. Both of which, of course, are wins.
Of course, just because starting the writing early is good, doesn’t make the process any easier. In fact, sometimes I think it makes things harder. If I had my experimental results, I could start by writing that section, and then work the rest of the paper around it. Instead, I have write my paper with imaginary “good” results, but be ready to tweak and re-organize the paper once I get final results.
But, as stressful as it is to work on a paper, it’s also a motivator and a sign of encouragement. It means your getting close to having publishable results. Which equals progress. So I guess I can’t really complain as I add another ball to my juggling act.