STATISTICS · MACHINE LEARNING · SOCIOLOGY
STATISTICS · MACHINE LEARNING · SOCIOLOGY
I want to start with a disclaimer. I am an empiricist and a social scientist. So I’m very sensitive to survivorship bias, and I tend to think the best way to figure out what will work for you is to look through a large repository of past successful examples. Conveniently, Alex Lang has compiled and hosts an incredible database of past successful (and unsuccessful) NSF GRFP statements. I used this resource extensively when drafting my application, and I would really suggest you do the same. In the interests of further discouraging you from continuing here are the links to my statement, proposal, and reviews, conveniently at the top of the post, so you don’t even need to scroll past all the words to find them!
Ok, so now that I’ve properly indicated that I don’t think you should listen to me, I have two pieces of advice for writing an NSF GRFP application.
First, this is almost surely the longest application you will write. The personal statement is quite long; mine is about 3000 words. I would take the opportunity to write a reasonably comprehensive narrative of your “intellectual journey” — I know that sounds insanely corny but here we are. My most generic advice for applying to anything is that you should be aiming to tell a story that unites as many of the things you’ve done, courses, research, extracurriculars, as possible into a coherent story about who you are and what you’re planning to do next. The story you tell should make clear why your plan going forward makes sense. Now you may be thinking “but I’m not sure I know what I want to do, how can I weave a narrative from the random shit I’ve done since freshman year that leads inevitably in a conclusion I’m not even confident in!”
And you know, that’s fair.
For some people it will be easy: if you’ve been dreaming of becoming a doctor since age 12 and you’ve done a standard pre-med track with a lot of bio research your story is probably pretty straightforward and things will just fall into place. If, on the other hand, you started thinking you’d work on engineering biological systems and ended up writing a thesis summarizing 350 years of legal history you might have a harder time (I say from experience).
So let’s assume you’re in the trickier second category. If you’re applying for the GRFP you are at least considering grad school, and you should probably know what department you want to end up in (although I didn’t). Here’s what I’d do: I would find some general area that you’d be happy working on that will let you tell as “pretty” a stray as you can about the classes/research you’ve done so far. If you’ve exclusively worked on plant biology but think you might want to pivot to human anatomy your GRFP application is probably not the place to go into that. No one is going to hold you to this, and its really important you write the strongest application you can. That means picking an area you can write a strong story for your personal statement and a well-informed proposal. If you read my application I think it will be clear that I was really looking for something, anything, that would tie together my several really distinct research projects.
The narrative I’m describing is really connective tissue though; the meat of your GRFP personal statement should be as detailed as possible descriptions of what you specifically did in various relevant course/research/extracurricular experiences. If you won any awards or published anything or are in process of publishing anything I would highlight that in bold (again see my statement).
I will also say that once you’ve written this for your GRFP it will be incredibly useful for your other applications. All my PhD personal statements were essentially cut down versions of my GRFP application.
Okay, on to the second piece of advice. The proposal is important, and you should pick a topic that you feel comfortable has not already been done in the literature and that you can write confidently about. If you have a research mentor I would talk to them about this extensively, and have them read your proposal and give comments. That is far and away the best way to ensure its cogent and makes a meaningful contribution. I will also say that like the personal statement writing a great proposal should come before writing a proposal that you actually plan to carry out, especially if you are a senior in undergrad (the whole process is actually easier once you’re in grad school, so take heart in that!). The NSF GRFP wants to invest in people more than projects; many people say this and it’s really worth taking it to heart. Your proposal should be the most novel research idea you can write about in a technical way for an audience who will be able to evaluate it on the merits. That will not always match with the research you really want to do over the course of your PhD, and that’s ok.
That’s it! Good luck! Feel free to email me if you think I can be helpful, though I’ll note if you’re not in the social sciences, CS or genetics I think I’m unlikely to be able to give helpful comments beyond those I make above. I’m especially interested in making myself available to people from underrepresented groups, non-men, and first gen people, so if any you resonate with any of those descriptions and would like more custom-tailored advice, definitely reach out me at njwfish [ at ] gmail.com.