NSF data management plans

Jul 20 2011 Published by under [Education&Careers]

Roughly a year ago the NSF instituted a requirement for a Data Management Plan to be included in every research proposal. Fresh on the heels of the Postdoc Mentoring Plan, this was yet another document outside the proposal narrative to think about. A few universities stepped up with some resources right away (summarized here at Book of Trogool) and the rest of us looked around to try and figure out what this was all about. BIO recently came out with some guidlines for proposal writers (PDF here) but the DMP is new to many reviewers.

Two rounds have passed and reviewers are starting to figure out the difference between a good and bad DMP. Much like the postdoc plan before it, it has taken some time for the expectations to settle in. So what do we know so far about the types of things reviewers are looking for? Here are 5 things that have surfaced in my experience as a writer and reviewer of proposals.

1) Samples. If you work in an area where collections are made in the field or lab, will there be representative collections available to others? If the work involves DNA or RNA, will the next scientist be able to request the exact DNA you used or even some material from the exact organism you extracted it from?

2) Distribution. Will there be publicly available off-site storage for key samples? If so, where? A museum? Herbarium? Repository? Secret sub-subterranean liar?

3) Raw Data. So where DO those lab books end up? How about digital data like gel or blot photos and sequences? Will they be available? Are they stored in a safe way?

4) Back up. Related to point 3, what is your data back-up strategy? Are you relying on a 5 inch floppy disk to keep your data safe or do you have a more comprehensive plan? Is there off-site storage or would a fire wipe out the whole project and all unpublished data?

5) Publication. Are you making your publications available by paying for open access or choosing open access journals? How broadly disseminated is your science?

Other fields likely have additional or different requirements than BIO, but these are the major ones that I have come across based on the proposals I see. Although the requirements are still in flux, reviewers are narrowing in on the expectations and you are better off including too much rather than not enough. Much like the PDM plan before it, it is unlikely that a mediocre DMP will have an enormous effect on where the proposal ranks, but POs are asking panels to comment on the DMP specifically in the summary statement, so there is a clear record of whether you are doing a decent job on it. Likely in this round or next, this will be a more critical piece of the puzzle and you'll get the hammer if you slap something together.

5 responses so far

  • Thanks for posting this! I had my suspicions that this was what they were looking for, but very little information has been given about the DMP so far. I think the DMP is a good idea, it'll force people to start using the available technology and start to eliminate the inevitable loss of data that HD crashes and stolen computers used to cause.

  • CoR says:

    Nice! 🙂

  • FSGrad says:

    Old post and old thought, but I'm procrastinating...THANK YOU. This post and its links, those about BIs, and other grantwriting ones were critical as I was writing my DDIG. I'm now 15k richer, so it was reasonable advice.

  • proflikesubstance says:

    Congrats on the DDIG and glad this stuff helped.

Leave a Reply