Repost: I do my research backwards. How about you? August 18, 2010
Posted by mareserinitatis in career, research.trackback
This is a repost from my old blog discussing how I like to approach research, and now, my work. I have been having some thoughts about my approach versus others, and plan to expand on this further. Thus, for today, I am reposting it so that readers can easily refer back to this and/or have it fresh in their mind.
And now for something completely different…
When I was at SU, I took a class on research methods, and our prof gave us the slides for a talk given by Simon Peyton Jones.
There is a ton of great advice in there, but I especially like (love!) his second paper writing model:
Idea -> Write paper -> Do research
His reasons for this are (directly from his slides):
• Forces us to be clear, focused
• Crystallises what we don’t understand
• Opens the way to dialogue with others: reality check, critique, and collaboration
As a subset of the “focused” point, I think it saves time. That may sound strange, but when you write things up before you do them, you have records of what you did, where you are, facts you may need, etc. It’s all there, in one place, and you aren’t scrambling to find the parameters you used, trying to remember which references provided which info, or struggling with how to apply the method you chose. It’s the difference between going on a trip with a road map and all your stops planned versus winging it and driving around every night trying to find a place to stay (let’s hope you don’t need a gas station, too!). The latter will give you quite an adventure, but I prefer to find my science adventures in other ways.
The second consideration is that I find it easier to sit down and write first. Jumping into a project without a roadmap is very intimidating for me. I often don’t know where to start and may spend a lot more time than I like flailing around. If I have it written down, I am able to bypass that “flailing at the intimidation” stage.
Finally, the best advantage is that, once you are done taking your data, it’s almost instantaneous gratification. Who really likes to take all that time doing their experiments and then face the ordeal of paper writing? The best part of doing it this way is that by finishing your experiment and computing the results, you feed them into the paper, and the paper is also done. Voilà!
This became abundantly clear to me when I was working on a project for a class. I wrote the paper first, a process which took me about 2 months. This took such a long time because I had to look up the data I was going to use and then compile it together. Once I had all the information compiled into the paper, I developed my computational model over a weekend.
So yeah, I’m hooked.
—–
One point that was made in the comments of the old post and which I feel the need to address is the notion that because I write things out ahead of time, I must also write my conclusions. I don’t usually do that, at least not completely. On the other hand, it is always a good idea to have an idea of what result you expect to see from your data, especially when doing simulation work. If your results differ significantly from what you expect, that’s usually a big warning sign to go back and check your procedures to see if you did something wrong. Perhaps it makes one more likely to say “data proves assertion is correct”, but I don’t think that following this method of doing research makes that any more likely to shove the paper out the door. More often, I think it makes me look at the results with a more critical eye.
I also blog at Engineer Blogs, home away from home to some of the best engineering blogs.
To some extent that’s what grant proposals make you do isn’t it? You propose an idea, write out specific aims, and predict what you are going to find. I often find grant proposals much more fun to read (partly because they have strict page limit, and somehow people put more work in them) than any manuscripts.
Ideally I would like to do the paper and the experiments at the same time, having having an outline of what I’m going to do. I feel that way you can be constantly reminded of what the big picture is, and not lose the focus when you are boggled with failed experiments :S
That’s pretty much what I do: have the paper there so that I am putting in the parameters from the paper, that way I can make sure I have no errors as far as setup goes.
I think you’re right about grant proposals, but not everyone writes up proposals for all or even some of the work they do. On the other hand, it sure makes you wonder if they would be more effective if they did. :-)