Comrade Physioprof raises a point I've always wondered about:
...this old saw has been brandished:
Innovation is, and always was, higher in smaller/younger labs[.]
There is never any fucken evidence marshalled in support of this claim.
I'll admit that smaller labs can be quick and nimble, especially in comparison to large research centers. On the other hand, larger groups often have the infrastructure and financial security to pursue high risk ideas without serious risk of being derailed.
Data would be nice. I would like to think that, were it available, someone would have cited it by now.
And while we're on this subject, given the kerfuffle over whether NIH should allow re-resubmissions for proposals, are there any data on how productive re-resubmissions ('A2') are versus resubmissions and original proposal awardees? (Although A2s might falsely appear to be more productive only because they've had longer to prepare preliminary work, which would then be credited to the A2 award at publication).
I have an anecdote going the other way: My most innovative and successful project was outside of my or my research group's established area of expertise and I doubt that it could ever have been funded. For example, there were review papers stating that what we were trying to do was impossible. My impression is that the project could happen without explicit funding, without any institutional obstacles, only because I'm in a big lab.
I'll agree with Morgan Price. I suspect young labs are more *productive* - the young investigator was hired because their research plan seemed like it would produce something useful (and was mainstream/safe enough to convince a majority of the department). The young investigator is also under the gun to produce something from their limited funding, so they're less likely to wander away and take a risk on something that might not work.
You want the real crazy stuff, you're better off in an established lab where the failure of any given project isn't such a big deal. Not all established labs bother to bring the crazy, but some do.
Of course, I don't have any data either - it would be quite interesting if such data existed.
I think large labs can be structured to be both nimble an innovative. It depends on the decision structure and hierarchy. I would bet the more top-down the lab, the slower and less innovative it is. The reason being all decisions have to go through many layers, and often the same layers. A lab with a more modular structure where groups can act semi-independently, but do come together to talk shop on a regular basis, might be better.
I like smaller labs, but that is just a personal preference.
There was something in Nature last year about this. Jeremy Berg from NIGMS did some analysis of "scientific productivity of nearly 3,000 researchers who were funded by grants from his institute in 2006".