Research proposals of those who win NSF Fellowships (vs those who don't)?

[InquilineKea]

I just looked at the research proposals of a few people who failed to win the NSF Fellowship, and they actually scare me a bit.

I know that the NSF cares about Broader Impact a lot, and that is where I might actually be strongest at. But the reviews look at intellectual merit first and foremost, and that's what scares me, since how can you convince the reviewers that your idea has more intellectual merit than a huge number of other very strong applicants?

Edited October 20, 2012 by InquilineKea

[Physwimic]

I think it is important to remember that the NSF is not looking for the most creative, important, and just jaw dropping research proposal. In fact, if you propose a project which might win you the next Nobel prize, you probably won't get an NSF. What the NSF is looking for in the proposal is that you can formulate a solid scientific proposal that is doable in ~3 years with a plausible hypothesis, valid methods of testing that hypothesis, back up plans for if/when things go wrong, and an understanding of why the research you are proposing is important. I guess what I'm saying is keep in mind that the NSF is more than just the basic ideas, its about convincing the committee that you can carry out those plans.

[InquilineKea]

Okay thanks. I just know that a number of unfunded proposals also have very good hypotheses, valid methods of testing hypotheses, and an understanding of research.

Regarding backup plans though, how many successful proposals actually include backup plans? I've read quite a few successful proposals by now and none of them have any backup plans, so I'm hesitant about including them until I see one.

[rising_star]

This is one of those times where your university can really help you. There's probably someone in the Graduate School/College office that oversees fellowships. That person likely has access to and experience with the entire NSF GRFP process. Contact that person and ask your questions, get feedback on your proposal, etc. It will serve you better than asking questions here, especially since you aren't even asking in the main GRFP thread that most people read.

[guttata]

My take, at least from the biological sciences point of view isn't so much that there is a true backup plan, but that you can still collect useful data from an experiment where your hypothesis ISN'T supported, i.e. negative data still tells you something novel. Not the same as writing out a backup plan so you won't see it explicitly stated in most cases, but very important and the hallmark of a well thought out and designed experiment.

Edited October 21, 2012 by guttata

[TakeruK]

How much space do you have? I haven't applied for NSF fellowships (being not-American and all), but I have had experience and success with the Canadian version of NSF. In these cases, I only got one single-spaced page to do it all! So there was really no room for explaining a "backup" plan. The way it works for us is that the evaluation committee are not going to be experts in your particular subfield. For example, the people that read my planetary science project proposals were scientists from all fields of Physics and Astronomy. The most important thing is to demonstrate that you are capable of planning a project (i.e. you're a good investment for their funding money) and also explain why your project will help the scientific community.

Physwimic gave really good advice. rising_star also mentions great resources. Another good resource is your own supervisor or other professors. If you are applying for funding for a project you're already working on, then you should definitely work on this proposal with your supervisor. If you don't have a project yet, then I would recommend that you sit down with a prof (or skype with a prof if necessary), explain that you are applying to NSF and talk about a project with them. These profs have experience on what is a viable 3-5 year project, and with their experience, they might be able to provide some extra insight.

Finally, I don't think fellowships necessarily goes to the most exciting projects. I think they want to see something that will both work and be useful to the community. Especially since the readers/evaluators may not be from your specific subfield, it's very important to motivate your project well!