Chris Blattman

PhD Students: Writing an NSF application

Share on email
Share on facebook
Share on twitter
Share on linkedin
Share on reddit

Judging from the average application I have seen (and their ugly cousin: the PhD application letter), more students could benefit from advice. I have six key pieces, primarily for students applying for the Graduate Research Fellowship.

This is the one where you have to be a US resident or citizen, and applying or just admitted to a PhD program. Not American? As it happens, I followed the same principles writing my application years ago to the Canadian equivalent, and I think the advice applies. Already at the dissertation stage? I’ll point out where the advice applies to dissertation improvement grants and even the big faculty grants too.

1. The most important thing is to show that you can ask and answer a clear research question. It means actually stating a clear question, showing how it fits into the literature and field, describing a coherent methodology, and hypothesizing results. This sounds obvious, but is still rare. Remember: the “S” in NSF stands for science. If needed, go back to your grade 9 description of the scientific method. It is a timeless guide.

2. The next most important thing: your research question should be compelling. By compelling I mean that someone in another area of your discipline can (a) understand it, and (b) find it interesting and important. Link it to big questions in the discipline, and to the public interest as well. Avoid esoteric topics, or at least describe them in a way your classmates in other fields can understand. This goes especially for basic and theoretical research, where the link is seldom obvious to non-specialists.

Note that neither #1 nor #2 were “Write what you think you will work on”. Chances are you don’t actually know what you want to work on yet. It will probably change five times between now and your dissertation. Writing something clear and compelling is more important. Therefore:

3. Write about the question and study where you can outline the clearest, most complete research program, one that will be the most compelling to a non-specialist.

This would never apply to a faculty research grant or a dissertation improvement grant, when sticking to your proposal is essential. For doctoral fellowships, the proposal is no more binding than a PhD application letter (which we all know bears little relationship with reality). We assume you don’t know what you want to work on. At most you will stick with your subfield. As in a PhD application letter, I think the NSF is interested in whether you are capable describe a coherent, interesting and ambitious research project. So do so.

4. Discuss your proposal with advisors (note the plural), several weeks or months in advance. Try to get at least one to read a draft and your revision. You should do the same with fellow students.

5. Revise, revise, revise. Drafts need to marinate in their mediocrity for a week before their flaws become apparent to you. Explaining your proposal out loud helps the process. This is a general principle that applies to all writing.

6. Please, do not describe a transformational personal experience. To be shunned at all costs: revelations in poor countries; childhood dreams; inspirational words from elderly people; books that changed your life. Even if true, it is tired and corny. Think of a personal statement as a CV in prose, one that links your experience to your ability to execute the project. Not a History channel biography.

Added or dissenting advice from other faculty and graduate students welcome.

26 Responses

  1. What about the judicious use of bold-face and italicized statements I see in NSFGRFP documents I’ve found from previous winners around the web. It seems strange, but is it actually helpful?

  2. Yes, I think this is primary. I meant to emphasize this in point #2, but perhaps it did not come out enough. Public interest and social is essential to broader impact, which is essential to distinguish your application.

  3. This seems like good, first-order advice. I’m an econ phd student; I’ve applied twice for NSF GRFP, once successfully. One thing I might add is that the NSF seems to care a lot about the “broader impact” of your proposal. I don’t totally understand what that means, but I interpret it as “what’s the social value of your project?” meaning why would non-economists care. It seems to me that the application should also make clear the answer to that question. I don’t know that this is the most important part of the application, and I’m curious if the instructions to reviewers mention it?

Comments are closed.