What if we threw a data-curation party and nobody came?

Dec 21 2010 Published by under Praxis, Research Data

So a lot of libraries and campus IT shops in the States are gearing up to deal with this whole NSF data-management plan thing. Websites are going up, would-be consultants are warming up their phones, plans are being planned (and sometimes even executed).

What if we build it and they don't come? Have we thought about this possibility?

I'm afraid my intrinsically Cassandraic nature only partly inspires these questions. We know pretty well from surveys and qualitative investigations (bug me for a bibliography if you like) that the average researcher hasn't a clue librarians can help her look after her research data. The said average researcher despises librarians, for that matter; she thinks that pukka information management can be taught to graduate students soup to nuts in a weeklong seminar, and she thinks that the real limiting skill for data management is deep disciplinary knowledge (which raises the question of why she typically leaves it to wet-behind-the-ears grad students, but…). The average researcher is dead wrong, of course (including about disciplinary knowledge being the sole limiter), but does she know that?

So let's imagine our old friend Dr. Helen Troia of the University of Achaea's Basketology department for a moment, faced with this new NSF requirement. Where will she go for help?

Well, she's probably going to call her NSF program officer first, an eminently reasonable thing to do. I hope the NSF has told its program officers to tell all the Dr. Troias of this world to look for help in their libraries—at least on their own campuses—but I'm not sanguine. What is clear, though, is that the NSF isn't going to manage Dr. Troia's data for her; at most, it'll give her a better idea of what she has to do to prove she's managing it wisely. So where does she go then?

She may also talk to her research-support office. Libraries: does your institution's research-support office know about your NSF-related activities? If it doesn't, better tell it. And she'll have a word with her local grant admin (she's lucky enough to have one) as well. Libraries: what do local grant administrators know about you?

If Dr. Troia's data are digital (not all data covered under the policy are, a point that bears re-emphasis), her next stop is likely to be her departmental IT talent. Libraries: if you are only partnering with campus IT, you may (depending on the way your campus is organized) be missing the boat. Find out where the people in small IT shops hang out, and reach out to them, too.

Now, departmental IT may well take on the job, but they are liable to do it ludicrously wrong. "Here, have some server storage space," they will say, ignoring questions of metadata, versioning, formats, organization, security, citability and other sharing issues, sustainability past grant expiration, and possibly even backup. I'm not sneering; with my own eyes I have seen a campuswide IT shop at a major research university, a shop that should assuredly know better, advertising unbacked-up storage as suitable for data-archiving needs. (No, I won't link. Yes, I am tempted to.) Again, it's a case of people not realizing what they don't know. NSF helper-elves need to be prepared to cope with that.

If departmental IT punts (as it likely should), then and only then will Dr. Troia approach campus IT. She will do so with fear and trepidation, as campus IT tends to be a Cthulhoid monstrosity, as fathomable as sunken Rl'yeh and approximately as helpful. Libraries: how are front-line tech-support finding out about your NSF-related services?

If none of the above people with whom Dr. Troia interacts points her toward the library, she won't come to the library. I wish that weren't so too. It's so. The inevitable corollary is that outreach efforts should not start with researchers. It should start with the layer of support and administrative staff with whom researchers regularly interact.

Even more cheerfully: none of this may work. We just don't know yet. We'll know much better in a year or so! Best have a plan for if it doesn't. Can you get a list of campus NSF awardees, to contact them individually? Do you have a few campus researchers who are willing to do projects with you? Can you get at the graduate students who are doing the real work?

Good luck. I think we'll all need it.

4 responses so far

  • Excellent point. Thinking along the same lines, I scheduled an appointment with our VP for Research the other day so I could walk him through the NSF-mandate assistance and services the library is ready to offer now and our plans to expand those services in the coming months.
    I'm hoping the Research Development office will be more than happy to have someone they can refer NSF-mandate questions to...but we'll see.

  • Gail says:

    Hi Dorothea,

    I agree with much of what you say - particularly working with other units on campus, especially research administration. All essential. Department-level IT was not something that occurred to me and that is definitely worth thinking about. I'm sure most of the small shops would be quite pleased to not have to take this on. I wonder if liaisons could help with that one.

    Nevertheless, whether "they" come or not, here's why I think it's worth all the party planning. Few faculty may think of libraries/ians in this capacity, but if we don't step up, that will never change. We should have informational sessions on the new NSF policy because (among other reasons) EVEN if no one shows up, it sends a message that we (and by we I mean a collaborative group representing multiple units, not just the library) are on the ball. We need to get that message out in as many ways as we can, if we truly aim to play a role.

    As an aside, I think people keep forgetting that the need may not be huge and it may not be immediate. Cornell, for example, averages around 150 NSF awards a year. Suppose we're 30% successful in getting those grants. That would mean about 500 proposals per year or ~10 a week if that effort is evenly distributed throughout the year (I have no idea if it is). I think we'd be off to a roaring start if 20% of the prospective PIs contacted us for help. So... maybe a couple of people will call us after the holidays? That's not all bad. For those of us who are still trying to decide exactly how to triage requests that buys us some time, but if we're somewhat prepared, we'll land on our feet.

    So you're right: None of this may work. But we won't know unless we try.


  • Dorothea says:

    Completely agreed, Gail. We can't not do this. We just need to do it as effectively as possible, and we need not to panic or give up if our initial efforts aren't hugely, wildly successful.

    Giving up is the possibility that worries me most -- though I'm gunshy, coming from an institutional-repository background.

  • Sam Searle says:

    Hi Dorothea - thanks for this timely reminder of the need for partnerships. There is no way librarians can do this on our own (especially if you are actually a single person in an institution looking after this stuff!). Supporting data management is an institutional responsibility - librarians, records managers and archivists, enterprise and faculty/school IT folk, specialist e-research / cyberinfrastructure teams , research office staff, faculty managers and grad schools all having a role to play in supporting researchers (and each other).