Over at
The Ambrosine Critique, I commented that I think a "project-oriented" approach is generally better than a "read-for-ideas" approach to getting going on dissertation research, even if the projects don't seem that great at first. I'd say "always be producing something" is a good motto--one that works for me, anyways. This means working on projects.
My comment generated a slew of questions about my experience and approach to doing research. I think all of these types of questions are important things to discuss with an adviser who I'm sure would have much better experience and advice than myself. Nevertheless, below is my response to the questions posed. Take them with a grain of salt; different strategies apply to different people; and the like.
What is the scope of “a project”? A vague idea? A particular hypothesis? A particular regression or data set?
I think of "a project" as a viable plan of action. As an applied micro researcher, I think this means an identification strategy and a potential data set.
How many projects do you have going at any one time?
I have generally tried to focus heavily on one or two at a time. Any more and I would be worried about making good progress on any of them. However, when there's a lull in my main project(s) because I'm waiting for feedback from my advisers, because I've hit a roadblock, or because I'm frustrated and need a break from it, I'll sometimes spend a day or half-day on another project that has been on the backburner.
How long before moving on to the next project?
Whenever the next project has a higher benefit than the opportunity cost :) Seriously, though, it depends. If a current project is "good enough" then I think you keep at it until it's done. The expected lifespan of a paper can be thought of like it is for people. A "good enough" paper in progress has a *much* higher probability of surviving to completion than a "good enough" paper that has yet to be born. This is another argument for getting started on projects even if you're not in love with the idea--if it survives after putting some work into it, it turns into a good project. However, sometimes you hit a dead end and sometimes substantially better paper ideas come along. My advisers have been a great resource for helping me prioritize research projects--they’re in a better position to judge what is viable/new/important/etc.
Have you ever given up on a project, or do they just become “on-hold” for a long time?
Yes and yes. When just starting to get serious about research, I replicated one of my favorite papers with the hope that some research ideas would jump out at me; I replicated it, no "good enough" ideas came, and I moved on. I also spent a good deal of time looking for betting market inefficiencies without finding anything too interesting--okay, that maybe that one was more for pleasure anyways. I've also started a couple of projects that I've since put on hold that I wouldn't mind coming back to at some point.
To what extent do you work on a research program? i.e. how related are your projects; do they each tell a small part of a bigger story? How important do you think this aspect is for the job market?
(The answer here will probably be sub-discipline specific… I suspect there’s more of a premium in public/labor/freakonomics to flex your technical muscles. But your ideas would form a nice baseline.)
At first, I don't think it's worth thinking about a "research program" at all. You only *really* need one good paper anyways. However, it's definitely good to have a well defined area of research when you go on the job market. With that said, one project often leads to other related projects so a cohesive dissertation can develop from a single idea. So I think this should be considered when deciding how to prioritize subsequent projects only after the first one has "taken off."