Thursday, June 10, 2010

Renewal

When submitting an NSF proposal on a research topic that is closely related to work done in a previous or soon-to-expire grant (i.e., essentially a continuation of a research project), the options are:

- Submit a complete, new proposal. Of course you need to be very clear in the project description why new work on the same or closely related topic is justified and compelling, but otherwise the new proposal is administratively distinct from the previous one. This is reviewed just the same as proposals on entirely new research.

- Submit an Accomplishment-Based Renewal (ABR) proposal consisting of up to 6 reprints of publications that resulted from the original project in the past 5 years (2 reprints may actually be preprints) and a summary (max 4 pages) about the new research proposed.

I like writing proposals, but I can see the appeal of assembling 6 reprints and preprints and sending them off with a short summary of the transformative new research (+ all the usual forms and information about "human resources" and so on). My impression, though, is that ABRs are quite rare in my field. Perhaps the program officers don't like to go this route because it's better to have a full-scale review to back up decisions. (?)

The obvious advantage of the ABR is, of course, the time it saves. You write 4 pages of new research rather than 15 pages, and you send a bunch of reprints and preprints instead of writing a Results of Prior NSF Support section in the project description of a 15-page proposal.

In the NSF Grant Proposal Guide, PIs are "encouraged" to discuss renewal proposals with the program officers and are "strongly urged" to discuss ABRs with the program in advance. I have a feeling that this is code for "No matter how great you and your research are, you definitely need to discuss this with a program officer before packaging up your awesome reprints and bypassing the full proposal route."

So, my fellow NSF supplicants, have you submitted a renewal proposal, how close was the new research to the old research, which route did you go, and how successful were you?

13 comments:

Comrade PhysioProf said...

In NIH land, renewal applications have a much better success rate than new applications. However, the process for application--and the length of the application--is exactly the same as for a new applications. The only difference between a renewal application and a new application is that in the former, you explicitly address the progress you made pursuing the specific aims of the prior incarnation of the grant, and the study section is given the summary statement from the review of the prior incarnation of the grant (but, interestingly, not the prior incarnation of the grant itself).

Renewal applications have a much better success rate than new applications mostly because study sections give them better scores than new applications. This is probably for a combination of two reasons: (1) If you were productive in the prior funding period, then it seems reasonable to give the benefit of the doubt in relation to common grant reviewer bugaboos; and (2) to give a bad score is perceived as cutting off an existing research program at the knees, rather than just declining to permit the initiation of a new one.

Anonymous said...

I'd never heard of ABRs as an NSF-wide possibility until here, demonstrating how rare they are in my field. Never seen one in any of the numerous panels I've sat on. Never heard of anyone submitting one.

Anonymous said...

I have never heard of an ABR and I don't fully understand the implications of submitting a new proposal that is a continuation of a previously funded one. This is something I need to figure out, and soon.

Greg said...

I just received word that my renewal was recommended for funding. This is my third renewal from an original CAREER proposal. All have been "new" proposals. I think if you want to try for an ABR proposal you've got to be someone with a big name at a top school. "New" proposals are the way to go for most of us, I think. I did get glowing reviews and was very happy about that. Unfortunately NSF budgets are way too small and they keep doling out small amounts to chemistry.

Odyssey said...

I'm on my third NSF grant. The second was very closely related to the first - in fact, as the panel summary noted, a "logical continuation" of it. The third is only vaguely related to the other two.

In all my reviewing for the NSF I have never seen an ABR. Nor have I heard of anyone submitting one. I've always wondered why they persist in listing it as an option.

a physicist said...

I haven't been in this position yet, the closest I came was after my CAREER grant, but I didn't think you could renew those so I submitted a regular proposal. I tried to make it clear that the regular proposal was intellectually a continuation of the CAREER grant. But it was written as a full stand-on-its-own grant, along the lines in your post.

But I have several NSF grants right now and I have been wondering about the exact same thing, if there's any chance I can get away with an ABR for any of them.

I did once referee an NSF proposal that was a full proposal, but had 10 out of 15 pages devoted to prior work. This was from a top-notch PI, too. NSF guidelines say no more than 5 pages on prior work. So I decided that was someone trying to do an informal ABR.

Anonymous said...

I am relatively new to the NSF proposal process, but didn't even know that an ABR proposal was an option!

Douglas Natelson said...

I have yet to see a part of NSF where ABRs are really allowed by program officers. My experience has been in DMR and ECCS, and in both cases there appear to be only "new" proposals.

M. S. AtKisson said...

In 9 years of doing nothing but grant applications, I've only seen one successful ABR, and it was invited by the program officer for a GK12 that had become a national model. I generally advise PIs to write a new proposal unless their program officer tells them, "Yes, and ABR would be a great idea in your case." In a new proposal, the PI can use that section called "Relationship to the PI's long-term goals" and "Results from prior NSF support" to place the new application in context with the prior and projected work.

The culture around renewals at NIH and NSF are very different, as Comrade PhysioProf said. However, I'm seeing something new with the new NIH application and review format with respect to renewals. Productivity in the prior funding period is no longer enough. The reviewers are being told to "review the grant in front of them" rather than the history of the project or PI. The PIs with solid programs that propose "drill down" experiments seem to be faring less well than before.

John Vidale said...

I've never submitted nor seen an ABR. I've heard talk of special creativity awards in other disciplines, which require less specific proposed work.

If someone is particularly worthy in my field, the program officer sometimes recommends submitting up to a 5-yr proposal rather than the usual 2-3 yr one.

EliRabett said...

Reviewers HATE ABRs, which is a great reason to avoid them. When they were first introduced, the response in the letter reviews was so negative that the sensible never tried.

Anonymous said...

I was on an NSF review panel with an ABR, several years ago.
I had never heard of them before that.
The famous and prolific scientist got it.
I have always thought I would like to try one - for the reasons state - but have not done so yet.

Anonymous said...

My advisor has been in charge of a (successful and prolific) multi-decade long study. My understanding is that much of the 30-year period was funded with ABRs. This is in ecology on a sexy topic that helps NSF get funding from congress. The last go-around, however, the NSF program said the standards for ABRs had recently changed and the program officer suggested my advisor write a new grant proposal instead of an ABR. So ABRs may be going extinct...