How to Run a Research Group

First point: I don’t know how to run a research group, and I have no experience in doing so.

Second point: That being said, something that I read recently has been making me think about this particular topic. I found it on the blog of Matt Might, a CS professor at Utah who had 15 seconds of webfame for “The Illustrated Guide to a Ph.D.” In an list of academic productivity tips, he mentions

Running a research group is a lot like running as small business. Make it real by branding your research group: give it a name. (Like U Combinator! [His group. — PKGW]) If you’re working a project, give the project a product name or a code name.

What got my attention wasn’t the particular “give it a name” idea or “small business” analogy, but the broader point. A research group is an organization. Not a big one, but one nonetheless. I generally hate it when people do the following, but it seems helpful here: let’s look up a definition of “organization”! Wikipedia:

An organization … is a social arrangement which pursues collective goals, controls its own performance, and has a boundary separating it from its environment.

The academic research groups with which I’m familiar tend not to be so good at the “boundary” part — they generally don’t work very hard at establishing an organizational identity. (Obviously, I’m most familiar with my adviser’s group, but it’s not the only one I have in mind. I don’t know much about groups outside of astronomy.) I think this is too bad: from what I’ve seen, people are a lot more enthusiastic about their work when they feel like they’re part of a team and that everyone’s pulling in the same direction. It also makes a big difference when everyone knows what their “collective goals” are. (Even if, I would say, they don’t necessarily agree with all of them.)

The other thing that I notice is that research groups tend not to have a lot of the infrastructure that one associates with organizations. No, you don’t need to write up an org chart for a group of seven people, but some things genuinely are helpful — particularly, I’d say, communications tools. I get the impression that many groups don’t even set up a shared email list, which amazes me given that in the year 2011 doing so is utterly trivial. Besides being practical, shared spaces like these help build up that sense of organizational identity.

Lately I’ve been thinking about how I’d try to do things if I was in charge. A standalone domain name and a website with a prominent list of members are, I think, mandatory. So’s an email list. A table of group publications seems natural. This may be a sickeningly corporate way to put it, but end-of-semester parties would be a good way to celebrate everyone’s accomplishments over the past six months. And speaking of sickeningly corporate, I even think it’d be great for groups to come up with mission statements and have lists of core values.

We do great science by:

  • Choosing the right problem
  • Using (or building) the right tool for the job
  • Keeping things specific, and
  • Knowing when to say “good enough”.

Doesn’t that make you want to go out and do some kickass research right now?

As I mentioned at the top of this post, I don’t actually have any practical experience in this area, so maybe I should pay more attention to the fact that I don’t see many of these ideas being implemented in practice. But astronomy does tend to do things differently — I’d be curious about what happens in other disciplines.