Friday, December 01, 2006


I have been proposal-writing lately. Mostly I have been working on the thorough revamping of a proposal that wasn't funded in a previous submission -- 'wasn't funded' sounds slightly better than 'was rejected', does it not? In the intervening time, I have been working on the project to acquire an awesome set of preliminary data because one of the main criticisms of the first proposal was that I didn't have enough preliminary data. This brings up two issues:

1. How much preliminary data is enough? It's a well-known conundrum of proposals that you have to do some of the research before you can get funded to do the research. informFortunately, many universities recognize this and provide small grants for pilot projects to acquire the necessary data to write a big proposal. For some projects, however, it can be difficult to know how much preliminary data you need before the proposal will fly.

In the past year or so, I actually did a lot of the research that I proposed in the earlier submission, and published 2 papers on this work. There are still a lot of interesting things to do, and the pilot study has resulted in some new and exciting directions for the work. I hope that my preliminary work has strengthened my new proposal by showing that the proposed work is feasible, and that the project is very cool.

BUT, I have had reviews for other proposals say that I had "too much" preliminary data. Figuring out what is too much and what is not enough and what is just right is a moving target because every reviewer is different. I've had other reviews by reviewers who basically wanted see all the results before they would believe the work was possible. This leads me to my next point:

2. When submitting a new, revised proposal for a previously rejected proposal, how much attention should you pay to the previous reviews? There is no one answer to that of course, but I will say that I don't think there is much point to including a 'response to reviewer comments' section in a resubmission. The worst of these sound defensive, and even reasonable ones typically focus on details and don't impress anyone. I think it's better to revise the proposal in the most compelling way possible and let it be reviewed as it is, without reference to previous incarnations. I am sure there are divergent opinions on that issue.

I know that in some countries, the project director can write a rebuttal to reviewer comments. I rather like that idea, but it's probably not practical for a system the size of the U.S. NSF. Lacking such a system, I think it's a waste of precious proposal space to write a rebuttal.


DancingFish said...

I am defending my PhD proposal next week and am already starting to work up ths years round of small grants from different societies. Most of these expect ans NSF-like proposal and should include some preliminary data. I have no preliminary data at all. All my work is field based and impossible to do at my home university. I feel this will hurt me (my field is mostly laboratory work) in grant submissions but don't want to wait another year- I might be completely done with the research by that time.
Do you think that the not enough/too much question changes not only with the reviewer but also they type of project, like in my situation?

Ms.PhD said...

Nobody expects PhD students to have much in the way of preliminary data. That is a different kind of proposal than we're talking about here. PhD projects are reviewed mostly on the "quality of the candidate" (e.g. your CV, what school you're at, who your boss is) and whether the project sounds like a good idea.

I can't say that I know for sure (by a long shot), but my feeling is that you need at least enough preliminary data to show that the most essential and risky aspects are likely to work in your hands.

E.g. in my field that means showing you have made the necessary reagents, that they detect what you say they detect or inhibit what you say they inhibit. And, if you're proposing to develop a new assay to measure something no one has ever measured before, you should try to at least show that your approach works on control samples. If you can't do that, try to show that something similar to the riskiest part of your approach works on something similar to what you're proposing to do.

You get the idea.

I think I got docked some points on my last proposal because I took it a little too literally when people told me I needed a whole manuscript worth of data as preliminary results. The reviewers made a comment that basically said I need to publish this stuff to gain credibility.

I think at some point a lot of preliminary data that hasn't already been peer-reviewed is too much work for the committee to deal with. I've seen the same thing with thesis committees- they don't really want to have to critically evaluate a whole lot of results that haven't already been peer-reviewed.

Re: rebuttals, though, that's a weird issue. I think it's important to have a dialogue and I really appreciate that about our publishing system, however slow and fatally flawed it might be. Grants are far in between resubmissions, though, that I get your point. It might be different people on the committee, and then having extensive rebuttals just influences the new reviewers negatively.

One thing I know some people do is try to send the revised proposal to a different committee. That may require a serious change of spin, though.

Female Science Professor said...

Sometimes the exact same proposal gets one review that says there isn't enough preliminary data and another that says there is too much. That's why there's no obvious answer to the question. As you say, it's best to get enough to show that the riskier aspects are doable.
Dancingfish - Ms. Ph.D. is right - I don't think you need to worry about this (yet).

Ianqui said...

FSP: How likely is it that any of the previous reviewers will see your new proposal? If the answer is not likely, then yeah, I don't think you need to worry about rebutting or addressing previous comments too much. But if the program director will tell you that s/he intends to get some of the same reviewers, then it might make more sense, right?

Female Science Professor said...

For a resubmission, some of the reviewers will be the same and some won't. When I see a proposal for the second (or third) time, I've never been impressed by a 'response to reviewers' section. I'd rather just see a great proposal. If I'm seeing a resubmitted proposal for the first time, I find those response sections even more irrelevant.

Anonymous said...

"When submitting a new, revised proposal for a previously rejected proposal, how much attention should you pay to the previous reviews?"

For NIH grant proposals, you are required to address the criticisms from previous review when resubmitting an unfunded proposal. And one of the explicit criteria for review of the resubmitted proposal is the adequacy of addressing the previous review. And my own experience has been that the reviewers of the resubmission take this *very* seriously.