What Your Mentor Didn't Tell You
Everyone tells you to begin writing your dissertation early, well before you finish your research and plan your defense.
For some, this may be possible. For others, however, I think this bit of advice is not necessarily clear and easy to follow.
If your project is like most, it is a living, breathing creature. You are not merely checking "to do's" off of your proposal; you are collecting data and changing your ideas based on those data and exploring new paths as the literature on your topic grows. For some, projects will change completely. For others, you may be scooped or your hypotheses may have to change with some new discovery occurring halfway across the world.
Thesis research is a dynamic process.
This is why writing your dissertation "early" is not always easy to do.
I don't want to hear the peanut gallery whining about "starting with the introduction". Sure, you can write literature reviews on your topic. In fact, if you have the time and energy, a review is a good way to get out a publication. If you think about the purpose and nature of an introduction, however, you will begin to see how intimately it is linked to the meat of your thesis. Your introduction is, thus, a growing, living creature also.
Now, listen closely, because I'm going to tell you something you can do.
For those of you who present posters often (more than 2 times a year, including the first year of your research), you can tune out. The rest of you should stay on board.
Pick up a pencil and a piece of paper. For you computer junkies, paper is that thing that is usually white and flat, and it comes out of a printer. A pencil is something you put in your hand and make lines with.
Now, grab all of the data you've obtained in the last 6 months (or longer if you have data from over 6-months ago that have nothing written about them). Pull out the data that can fall under a single-hypothesis.
Now, write an abstract. I'm giving you a 300-word limit. You must have a rationale, methods, results, and conclusions/discussion section. Go online to a society (like the society for neuroscience) and look through their abstract databases for help, if you need to. It's sometimes helpful to have a model when you write (a model does not mean copying passages, just so we're clear, here). It doesn't matter if your data aren't finished or if you only have an N of one. Write about what you have. Do it now. Don't sit there and think, "I'll do it when I have time" or "I'll do that on Cinco de Mayo". Do it now. It's 300 words; you can do it. You are not too busy because you are reading my blog right now. You can't fool me.
Now, put the abstract in a file that you will not lose. Remember that you may need to pull it out in 5-years, so label it with care.
You now have a skeleton ready for dressing when you want to submit an abstract or write a paper. When you go to write your dissertation, you have an immediate outline to use for the chapter covering the data in that abstract. You can change the abstract as you go along in seconds. Quick, easy, done.
Thank me later. I accept all major credit cards.