Another reason data services need librarians

Some people watch football over Thanksgiving weekend; I get into discussions of disciplinary data regimes with fellow SciBling Christina and others on FriendFeed. Judge me if you must!

Another common truism in both the repository and data-management fields is that disciplinary affiliation accounts for a lot of the variation in observed researcher behavior. For once, I have no quarrel with the truism; it is unassailably the case. The wise data curator, then, knows some things about disciplinary practices going in.

But what things, exactly?

I don't believe that taxonomy exists yet; it'd be an awfully fruitful thing to research. Based on learning from Christina about chemistry and reading quite a few reports of late (including the one she references), here's my wild stab at a beginning:

  • Is research in this field collaborative or lone-wolf? In astronomy, everybody relies on everybody else's data; as data tend to be so expensive to collect and store, there's hardly any other way to work. In chemistry, lone-wolf labs tend to clutch data close.
  • What are the ties with industry like in this field? The closer those ties, the likelier researchers are to muzzle data, fearing that a scoop will mean monetary as well as career damage.
  • What are the ethical and legal constraints on data sharing? Sometimes these are obvious, as with human-subjects research. Sometimes they're less obvious but very, very salient.
  • Are existing players selling data? Which data? Self-explanatory, I hope.
  • How standardized are research practices in this field? How digital? Less important to data sharing per se than to the type, quality, and reusability of the data one is likely to encounter.

None of these factors is determinative, and all have counterexamples—even high-dollar industry doesn't always muzzle data, as genomics bears witness.

Here's the thing. I myself can answer the above questions for only a couple-three fields. If I'm responsible for a campus-wide data service, I need answers for every field the campus does research in!

Moreover, the answers to some of the above questions are ticklish and politically charged, not to mention that your average researcher will expect me to have those answers up-front. I don't want to question a researcher's favorite society's practices in a discussion with that researcher! Likely as not I'll lose the researcher's goodwill on the instant.

So until disciplines are clearly mapped out with respect to the above questions, or appropriate approaches determined for me by well-conducted research, I need a third party with a broad perspective to ask. That, to me, spells "librarian."


More like this

I years ago gave up on the possibility that anyone could know the varied practices that would affect data services across all research disciplines. Now, of a sudden, it seems possible. Who are we going to write a proposal to to get funding?