Data Tales from the Hills: the Request For Information (RFI)

At the latest #RDAP15, twitter was aflame with ideas. One was to chronicle our experiences with databrarianship. I’ll kick off our Data Tales series with a phenomenon that I hadn’t seen before: an additional Request for Information (RFI) from grant reviewers who want to award a researcher money, but can’t until the Data Management Plan (DMP) is in better shape.

This is how it goes down:

Random e-mail thread makes it to me: “Does our institution have any standard response to this?” from an unknown Sponsored Program Officer in a far-off College. I scroll down to an e-mail from the PI, “Please see attached”. I open an attachment and RED CAPITALS JUMP OUT.

“NOTE:  THIS IS NOT A GUARANTEE OF FUNDING. THIS IS ONLY A REQUEST FOR ADDITIONAL INFORMATION FOR PROGRAM REVIEW.” It goes on to detail six points of clarification, all focused on the data management plan. And there are 8 days to respond.

No one is in town (including me), so the ping-pong of e-mails begins, summarized below:

My response: No, we don’t have a standard response, since we don’t have standard data management capabilities. However, we do have some options. Would they care to use the DMPTool to address the RFI?

PI: delegates the response to a member of the research team at another institution.

Delegate: prefers e-mail consultation over the DMPTool.

Discussion of the points in the RFI ensues, but phrases like “It is primarily the responsibility of the institution that receives the grant to work with the PI to provide data access” and “verify that the institution will make the shared data available for at least 10 years” are making everyone nervous. Which institution? What solution?

Some of the material will be commercialized; all of the research involves human subjects. No IRB has yet been filed and how will they perceive any data sharing promises?

I point them toward phrases like “data sharing will be done in consultation with the IRB and Technology Commercialization Office” and indicate the various local policies that apply.

For data sharing, I lay out their options: A disciplinary repository (ICPSR) or our Institutional Repository (IR), which only accepts datasets on a case-by-case basis.

The researchers conclude that there isn’t enough time to get ICPSR on board and decide that they don’t care about the lack of DOI or that SPSS files aren’t supported in our repository. They compare their options but the other two institutions on the grant are not any better. The research group reviews the e-mail thread and give their thumbs up to depositing the data in our institutional repository.

I contact our IR manager and make sure the details line up. Green light: we can intake the SPSS file formats even though we don’t officially support them.

The entire consult took twenty-five e-mails in six days and involved six people at three institutions! We made the deadline, and the RFI responses weren’t half-bad.

Do I wish it were easier, that we had a robust data repository and standard language? Absolutely!

Do we have that yet? Unfortunately, no. Nor do many other institutions. Until we do, I’ll do my best to make sure the data management plan isn’t a reason to be denied funding.DMPAPPROVED1

Update: Since this episode, I’ve seen three grant review boards request more information about the data management plan prior to awarding funding, including the Department of Education, and two NSF directorates (Education and Human Resources, & Computer and Information Science and Engineering).

Is this a new phenomenon in the grant world or something just now making it to my office? It’s hard to tell, as reviewer comments are typically only revealed to the research team. However, if this is an increasing trend, then it’s an indicator that grant review boards are taking data management plans seriously.

– Amanda Rinehart

~~~~

Do you have a data tale to tell? Send it to Databrarians editor Amanda Rinehart (Rinehart.64@osu.edu). Please get permission from the researcher before sharing their story, or thoroughly de-identify: we want to learn and not to expose other learners! 

One Comment

  1. DK said:

    To me the lesson of this story is that researchers should take the DMP requirement seriously and seek consultation with institutional resources (data librarians, the staff of local IRs or the domain repositories that their planned data might be a good fit for) IN ADVANCE i.e., while they are actually preparing their research proposal and its accompanying DMP.

    December 8, 2016
    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *