Posted by: Jeremy Fox | April 27, 2012

Is ecology becoming too collaborative?

Is it just me, or is everyone overcommitting to too many collaborative projects lately? Everybody likes to collaborate these days. It’s easier to do these days, and the culture of ecology values it highly. And frankly, everyone sees it as a way to get their names on a paper while doing only part of the work. So we all agree to participate in lots of collaborations–to the point where we lack the time to really pursue any of them. Which, paradoxically, only encourages us to agree to more collaborations. After all, if your collaborators are moving slowly, you may as well start up some new collaborative project with someone else while you’re waiting for them to get their act together. And if you’re moving slowly, you may as well start up some new collaborative project with someone else while your other collaborators are waiting for you to get your act together.

Don’t get me wrong, I’m not saying we should all stop collaborating. But anyone else think this sort of dynamic is becoming more common?

About these ads

Responses

  1. Hmmm, this could easily be studied statistically. Fancy doing something together on it?

    • Wow, you win the thread with the first comment! ;-)

  2. I recognize that and I think there is a tendency for increasing numbers of multiauthor papers and also papers are having much more references cited (see Neff & Olden 2010 Bioscience). Of course having more references cited in a paper can have several explanations (e.g. more information is available now, or more self citations because there are several authors), however I think that this tendency is because the “publish and perish” rule is taking the academic world and a possible solution for having a reasonably scientific production is having more collaborative research, thus increasing the chances of producing more papers. I think in the current days high impact journals are publishing only global scale issues where simply your data set is not enough and you need something more, something that your colleague elsewhere got. So people wanting to publish big picture research must to collaborate.

  3. Speaking as someone who’s a part of a big collaborative project: I hope not! Although you don’t really present us with a down side in your post.

    Big collaborative projects help you produce more comprehensive work by using specialists from many fields, it helps you as a researcher since you begin to learn other scientific languages, as you noted, you can possibly get your name on more papers (although surely that’s only a cynic’s primary motivation), and you can do some really great science.

    The downside might be more problems with data and personnel management, slower times to publication. . . Maybe I don’t want to see the downside, but since you posed the question, I’ll leave it up to you!

    • Well, I think the downside might be slower collaborations on a per-collaboration basis. Imagine that in the limit everyone collaborates with everyone, and no one gets anything done because we’re all too overbooked to finish anything. ;-)

  4. It’s absolutely becoming more common and has been for at least a decade. I think the main driver is funding opportunities. Outside of NSERC DG, there are very few funding sources that fund a single P.I.. Especially for big $$$ there is typically a requirement for research groups and consortia. A recent funding opportunity I looked into requires international collaboration, including a mix of natural and social scientists. Although I am naturally a collaborative scientist and enjoy collaborating, I have certainly been a victim of its pitfalls (i.e., politics, people sitting on data/manuscripts etc.). With the current funding climate I imagine funding agencies will continue to want “more bang for their buck”, so collaborative research is likely here to stay and dominate the research landscape.

  5. I’ll jump out on a limb and say, “No.” Rather, I think we’re dreaming up and executing more and more projects that are too big for any single investigator. Thing of NutNet, NEON, projects involving the LTER network, any NCEAS working group, or pretty much any experimental (and often observational) ecology on a global scale. Collaboration is necessary, and these sorts of projects are becoming increasingly common – and increasingly necessary as many ecologists try and address problems at a very large scale.

    So, maybe the question isn’t “Is ecology becoming too collaborative?” but, “Is Ecology doing too much at large-scale integrative work?” Or something like that – you get my gist.

    • These are good points about the requirement for collaboration with increasing scale and complexity of research. While I agree that this explains some of the increase in collaborative research, I think a lot of the increase in collaborations is due to lower costs of collaboration nowadays (thanks to communication technology) and lower standards of authorship. One study (Wuchty et al 2007) analyzed the number of authors in diverse fields of research from 1955-2000. They found the number of authors per paper has increased not only in Big Science fields but also fields in which scale and complexity have not noticeably increased over recent decades. This suggests there are reasons beyond the scale and complexity of research that explain the increasing prevalence of collaboration.

      It seems to me that as our job prospects depend more and more on citations there is increasing pressure (real or perceived) from researchers to be added as authors to papers. That fuzzy line distinguishing an acknowledgement from authorship has probably shifted to more liberally grant authorship. The benefits of this are experienced directly by the researchers themselves, while the costs are a Tragedy of the Commons type experienced by the research community as others’ hard-earned authorships are diluted by frivolous authorships. Lax authorship likely pays off in our current system. I wonder if this trend would change if employers shifted their focus from publication-counting to one of those fancy versions of the H-index that partition credit to authors according to how many authors appear on a paper (e.g., Batista et al 2006)?

      REFERENCES:
      Batista, P.D. et al. Is it possible to compare researchers with different scientific interests? Scientometrics, Vol 68, No. 1 (2006), pp. 179-189.
      Wuchty, S., B. F. Jones, and B. Uzzi. 2007. The increasing dominance of teams in production of knowledge. Science 316:1036-1039.

      • Good points, which of course are related to Nick’s comments. How should we evaluate the contributions of individual authors to multi-author papers, especially ones with more than a few authors? Because we do often have to do this, for instance when deciding who to hire. One answer is to say that authorship lists can’t actually encode very much information about author contributions, and so the “author contribution statements” now required by some journals should become routine.

        Personally, I don’t find formal indices at all helpful, even if they attempt to divide credit for a multi-authored paper among all the authors. In the contexts in which I’ve needed to compare or evaluate people–hiring my own postdocs, sitting on faculty position search committees, evaluating grant applications, evaluating tenure and promotion applications–I’ve never felt a need for such indices. Yes, if you have a lot of applications for a job, there is a need for some quick and dirty filters that let you quickly weed out the many uncompetitive applicants. But frankly, that’s not a very difficult job, and there are all sorts of informal filters that do that job at least as well as formal indices. And in order to compare competitive job applicants, or evaluate people for tenure, or evaluate grant applications, there’s no substitute for looking at all the information you have, of which any citation index would only be a very small part. And that’s before you even getting into all the widely-discussed issues with interpreting those indices, most of which have nothing to do with apportioning credit for multi-authored papers.

        Sorry to get on a soapbox in response to your passing remark on citation indices, but it’s an issue that really bugs me. I actually think people overrate the extent to which hiring, grant application, and tenure and promotion decisions are based on just “counting papers” and “counting citations”, and so I think the whole search for better citation indices is misguided.

      • I agree with you, Jeremy, that indices are over-used, and I’m glad to hear they might not be used as much as I fear. I published a letter to the editor in Nature last year criticizing authors’ reliance on the Impact Factor as the only factor to consider in deciding to what journal they should submit their manuscripts. There are plenty of ways for journals to bias their impact factor that have nothing to do with improving their scientific quality. Impact factors and other indices are useful, but like you say they shouldn’t be the only (or even major) factor to consider when evaluating researchers or journals.

      • Oh, don’t get me wrong. While I don’t think that citation indices have as much influence as many seem to think on hiring, tenure, and grant decisions, it’s clear that journals are increasingly (and increasingly openly) doing things to chase high impact factor, and it’s increasingly common for authors to first submit their mss to high impact journals in the hopes of getting lucky.

  6. What counts more for tenure and promotion:

    1. One single authored paper?

    2. Two papers with two authors each?

    If the second counts more, there are gains from trade. “Let me put my name on your paper, and I will put your name on my paper”. You can even do a bit of work on each others’ papers, just to make the trade legit.

    • Good question Nick! To which I’d add: what counts more for getting hired, and getting an NSERC grant?

      And I’m not sure of the answer. Probably many folks would say (or think others would say) #2, all else being equal. But I think that’s true only to a point. If for instance you’ve co-authored a lot of papers, each with many authors, and you’re never the lead author, I think those who are evaluating you will start to wonder if you can think for yourself. Personally, I like to see someone who has a mix of sole-authored and collaborative work. I take it as a sign of someone who has their own ideas, but who can work both independently and collaboratively in pursuit of those ideas, and who is well-hedged against both setbacks in their own research and in their collaborations.

      I don’t know that “gains from trade” are much of an issue, not because sole-authored papers are more valued but because professional norms prevent such trades. I do think considerations of “what will get my name on more papers?” are one factor driving the increasingly collaborative pursuit of ecology, but so far that’s manifested as more people trying to pursue more legitimate collaborations, rather than as trade in co-authorships. I don’t know if professional norms will start to break down at some point, but I sure hope not.

  7. Jeremy: OK. Forget about the “gains from trade” model. Instead, lets think about a “Darwinian selection” model.

    Some young grad students follow strategy 1 (sole-authored papers). Other young grad students follow strategy 2 (co-authored papers). If two halves beat one whole in a tenure or promotion committee, strategy 1 will die out, because they are less likely to survive hiring and tenure, and because they are less likely to get promoted and get NSERC grants and breed more grad students like themselves.

    You get exactly the same predictions using the “Darwinian selection” model as I get with the “gains from trade” model. It just takes a few generations longer.

    • Hi Nick,

      Sure–as long as two halves beat one whole, and there’s no possibility for mixed strategies. In fact, I think a mixed strategy often has the highest payoff, all else being equal. I’m also not so sure that two halves always beat one whole even if only pure strategies are allowed.

  8. I’m looking for collaborators on my GUTPEC model (described here previously) if anyone’s interested. There will undoubtedly be a whole string of papers and much fame to be had.

  9. Seriously though, a lot of these so-called collaborations are a bit of a joke in a number of respects. You’ve got 10 authors on a paper you’re reading but have a specific methodological question, how are you supposed to know who to ask? I have my serious doubts that in a lot of these cases the lead author even *knows* who did exactly what. And that’s a big problem in my book.

    • Well, I would certainly hope that the lead author would know who did what! So yeah, if I had a question I’d just go to the lead author fully expecting them to either answer it or tell me which co-author can.

      • When it’s 2 to say 5 authors, they probably often (but not always) do, but when it gets toward double digits, I have my considered doubts.

      • Author contribution statements are one way to address the concern you raise.

  10. Not to mention the whole issue of there being a potential reluctance to criticize the work your co-authors have done–I think you raised this point previously didn’t you? When people tend to work in smaller groups there is arguably a greater tendency to “let ‘er fly” with the criticisms, and that I think, is a very good and necessary thing. This is NOT a trivial consideration methinks.

    • Yes, although as commenters on that old post noted, different groups of collaborators can interact in very different ways. Some commenters had been part of NCEAS working groups in which folks spent a lot of time (productively) arguing and pushing one another. Some people just like a good argument. Or so I’m told; I totally don’t know anyone like that. Nope, definitely can’t think of anyone like that at all…What? Why are you looking at me like that? ;-)

      The other thing is that one often ends up collaborating only with people who share one’s basic point of view. People don’t just collaborate with randomly-chosen collaborators. For instance, in the CIEE plankton dynamics group I organized, all the invitees had sufficiently similar “world views” that we weren’t going to have any really serious arguments. Unless the whole purpose of the collaboration is to bring together people with opposing points of view, of course, in which case there are obviously no worries about people pulling their punches.

      • Yeah, good points. Probably hard to generalize too much about the pros and cons of author group size–lots of particulars at play surely. I think that if you look for wide and open minded thinkers that things tend to work out positively.

        However, it’s easier to simply banish anyone that doesn’t agree with my viewpoint, so that’s the route I prefer.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Categories

Follow

Get every new post delivered to your Inbox.

Join 4,148 other followers

%d bloggers like this: