April 23, 2007

The Rise of Team Science

Posted by Eric at 10:41 am | Category: Literature, Science

An article published recently in Science Express focuses on the growing trend of team-based research in science and engineering research, and the decline of the individual. This has all been known for quite some time, especially with the rise of massive collaborations a la the Human Genome Project and large particle accelerators, and also because of increasing interdisciplinary research that requires teams of specialists. The trend, however, is apparently also increasing in the social sciences and mathematics, where lab-work isn’t always as critical to the success of the project and where one might think that having too many people might even hinder good work.

The most interesting aspect of the paper was their analysis of the quality of team-based research. They showed was that team-authored papers in science are being cited almost twice as much as solo-authored papers are, even with self-citations removed.

One could also consider that, although solo-authored papers would be cited less on average, they might make the greater portion of the hugely influential papers. In the past, apparently solo-authored papers were more likely than team-authored papers to gain the highest numbers of citations, but now, team-authored papers are more likely to gain a high number of citations.

It’s clear from the paper that in all disciplines, from science and engineering to mathematics and the social sciences and even to the arts and humanities, team-based research is coming to dominate the solo researcher. The image of the lone-genius is slowly dying, I think. Maybe in the future students will learn about the Grubbs-Evans-MacMillon-Sharpless (GEMS) catalyst, or the Schreiber-Lander-Maniatis-Weissman-Walsh-Eggan-Melton (SLM-WWEM) embryonic stem cell protein complex!

Follow the comments thread by the RSS 2.0 feed. You can leave a response, or trackback from your own site.

2 Responses to “The Rise of Team Science”

  1. Apollo Says:
    April 23rd, 2007 at 2:38 pm

    I’m sure funding is part of it: both PIs I have worked with were applying for group grants with three or four other researchers when I worked with them. Perhaps there is the growing assumption that funding a group of researchers will undoubtedly produce some results (even if they aren’t eye-popping breakthroughs) while funding a single researcher might not yield anything? If anything, one would hope that there’s more accountability with a group and less of chance of data being fabricated (i.e. it’s easy to lie, but somewhat harder to get other people to lie with you - unless you’re dead sexy).

  2. Eric Says:
    April 23rd, 2007 at 5:59 pm

    Risk-averse funding is definitely contributing to the phenomenon, and people are definitely collaborating more because of how thinly funding gets spread sometimes (and concentrated at other times). I know in high-energy physics, money is definitely a factor, because any experiment requires such a huge amount of funding that only a band of people who pool their grant money can get anything done. Things like the HapMap project and various genome projects can only happen with lots of collaboration, too, just because of the sheer amount of labor needed.

    Maybe data fabrication is a little harder in a group, but not by much, I’d say. Usually, there’s a division of labor with teams, and so if one person in the team is lying to another, the rest of the team probably won’t be able to tell easily. It sucks, and it’ll eventually take down the entire group, but it happens. Mistakes happen in group environments, too, like the Chang protein structure retractions, and those get through enough times that I think group dynamics doesn’t help as much as one would hope.

    At some point, pooling more money and people won’t help that much for risk and such, I think. Otherwise you’d always have 100 person grants! And really, funding five labs to work on one project isn’t as good as funding three labs to work on that project, if the research isn’t labor intensive, because concentrating the money allows one to do things (like buy instruments) that might not be as feasible with five groups and the same amount of funding. I think that a large part of the growth is just the need for more labor and expertise in interdisciplinary work, the growth of huge-scale projects, and for resources that particular labs have that other labs need (antibodies, machines and instruments, algorithms, etc.)

Leave a Reply