This is my second post about the NSF GRFP. My previous one was about the personal statement. If you plan to apply for an NSF GRFP, you should definitely read that one first, and then read this post about the research plan. If you don’t plan to apply to the NSF GRFP, there is no need to read either post.
After the personal statement, the research plan is the next-most important component of a GRFP application. It is where you describe what research you plan to carry out. And you only have two pages to do so. This means you can’t actually write a detailed research plan. What you have to write is an outline of a plan. I think of the research plan as an extended summary page. If you’re familiar with the standard NIH proposal structure,1 think of the Specific Aims page, and just expand it a little bit. There you go. You’ve got your NSF GRFP research plan. (But don’t forget the Broader Impacts section.)
There’s a lot to be said about how to write a research plan, and I can’t possibly fit everything into a single post. So I’ll start by linking to a few other proposal-writing resources, and then I’ll provide suggestions that are specific to the NSF GRFP. First, I suggest you go and read this old blog post of mine about fine-grained sectioning in proposals. Then, you may find this blog post about the critical need helpful. Finally, here is a handout from a class I taught some years ago. The handout has various pointers about writing grant proposals. Most importantly though, on the second page, it has a sentence-by-sentence outline of a proposal summary page. Just follow the outline and your research plan will write itself.2
Write for a broad audience
A common mistake I frequently see students make is to assume everybody knows about their research area. Students may have spent years in a lab where everybody talks about the same problems day-in day-out and now they think anybody who is broadly in the same field will be intimately familiar with the specific questions and issues they deal with every day. Nothing could be further from the truth. It’s best you assume nobody knows anything about your research. Yes, the reviewers are all professors in your area. And no, they have never heard of the system you’re studying.
To give a concrete example, in biology, students working on computation frequently misjudge what computational methods the average experimental biologist is familiar with. As a consequence, they write proposals that most reviewers can’t understand. And the reverse is also true. Experimental students may assume certain approaches are well known when maybe only ten labs in the world do these particular experiments.
So, do your best to write for a broad audience. But also, be aware you may still misjudge what is and isn’t broadly known. As a simple test to assess how comprehensible your writing is, show your proposal to a fellow student in your cohort but not in your lab. If they don’t understand your research plan chances are the reviewers won’t understand it either.
Stay away from biomedical research
There’s an unwritten rule that the NSF doesn’t like to fund anything that is clearly in NIH territory. The NIH supports research related to human disease. Therefore, the NSF generally stays away from human disease. The reasoning is that the NIH has a budget that is five times bigger than the NSF budget, and yet the NSF is tasked with funding all of science whereas the NIH only funds research into human disease. So that’s one area the NSF doesn’t see the need to spend its scarce resources on.
If you’re a physicist, a chemist, a computer scientist, or an engineer, there is little risk that you’ll accidentally propose something that looks like NIH material. But, if you’re a biologist, this is a realistic concern, since the majority of biological research in the US is funded by the NIH. If you’re an undergraduate or first-year graduate student doing research in a biological lab in the US there is a good chance that your lab is NIH funded and so you’ll just instinctively absorb NIH language and priorities. If this is the case for you, I encourage you to deliberately stay away from NIH language and from human disease.3
Don’t worry about committing to a specific research program
Receiving a GRFP award does not obligate you to do the exact research you have proposed. It’s easy to see why this should be the case: The vast majority of students who are applying are either not yet in graduate school or may be in graduate school but are in a rotation system and haven’t yet chosen a permanent lab. In both of these cases, students may end up in a situation where they literally cannot do the research they have proposed. For example, if you propose to do research on penguins, and then you don’t get into the penguin lab or even the school where the penguin lab is and instead end up in a butterfly lab, you’ll have to do butterfly research. But even if you do end up in the penguin lab, by the time you’re situated and the GRFP has started almost a year will have passed from when you applied and the specific research you suggested may no longer be relevant, or interesting, or the most impactful way for you to spend your time. Or, you end up having a long conversation with your advisor and they explain to you why maybe you were a bit naive when you wrote your research plan and some parts of it are infeasible, too costly, or uninteresting. In all of these cases, the right thing to do is go after the research that makes the most sense, not the exact research you proposed.
What you are writing in the research plan is an outline of what a reasonable PhD project could look like for you. You’re not writing a contract in blood that you’ll be tied to for the entire three years of GRFP support. Changes in research plans are normal and occur all the time. Changes in advisors are normal and do occur. Even changes in graduate program or university are normal and occur from time to time. You can be awarded an NSF GRFP and still make any of these changes in your PhD research if you want to.
Write a research plan about the research you know best
It is a good idea to be somewhat strategic about the research you propose, in particular if you haven’t yet joined a permanent lab. Write about the research you are most familiar with, even if it’s not necessarily exactly the project you want to do for your PhD.
As an example, let’s assume there are two labs you might want to join, one you’d be very excited about but it’s a little further thematically from research you have already done, and another one that is more aligned with your prior research experiences but it’s less exciting to you. In terms of the proposal writing, you will likely be better off writing a proposal related to the second option, even if you’re confident you’d join the first lab if both options were available to you. In proposal writing, lack of specificity or detail is the kiss of death. It’s unlikely you’ll be able to write a strong proposal about research you haven’t done previously and have no experience with.
Structure your research plan into well defined objectives
Your research plan needs an overarching theme or goal or research question, and then two or three well defined objectives you will pursue to achieve the overarching goal. The right number of objectives is critical. Three is ideal, and two is fine. More than three is too ambitious, and less than two4 means you haven’t actually developed a proper plan.
Each objective should have a descriptive title in bold and then about a paragraph of text describing what you will do. It can also help to add a hypothesis after the title and before the paragraph of text. To get an idea of what I mean, take a look at my old blog post on fine-grained sectioning in research proposals. You can also look at the first page of this old grant proposal I’ve sometimes used as an example in classes I’ve taught. But, note that both the blog post and the grant use the term “Aim.” Write “Objective” instead. Remember, you’re not applying to the NIH.
Address the broader impacts and the intellectual merit
In your research plan, you need a separate section called “Broader Impacts.” Realistically though, you don’t have sufficient space in your research plan to write much about broader impacts. My recommendation is to limit this section to a single paragraph, in which you summarize the broader impacts plan that you have spelled out in more detail in your personal statement. See my post on the personal statement for details.
And you also need a section on “Intellectual Merit.” My post on the personal statement addresses this also, so I’m not going to repeat myself here. Go read the prior post if you haven’t done so yet.
I realize there is probably no overlap between people familiar with the standard NIH proposal structure and applicants to the NSF GRFP. But, if you’re an applicant, maybe your adviser is familiar with how NIH proposals are structured. If they are, ask them about the Specific Aims page. Importantly, never use the word “Aim” in your NSF proposal. NIH funds Aims. NSF funds Objectives. DOD funds Tasks.
Assuming you have a good idea of what research to do. I can’t help with that. At least not in this post.
Have I already mentioned that you shouldn’t use the term “Specific Aims”? You’ll be proposing “Objectives.”
In case it’s not obvious, less than two means one. Because obviously you can’t have zero objectives. You need to do something.
Hi, thank you for these in-depth posts! If I DO want to ultimately do research on human diseases, are there similar NIH fellowships for funding PhD students? I know there’s an NIH F31, but that seems to be for already-enrolled PhD students (which I’m not). Trying to yoink funds from other scientists seems a bit disingenuous if I’m going the medicine route.