For those who don’t know, a PhD program is an opportunity where you are paid (tuition & fees covered + living stipend) to do research and drive knowledge discovery. However, money doesn’t grow on trees, so how are we going to get this bread? 💰

One way is by securing a fellowship that will fund your PhD, and the Graduate Research Fellowship Program (GRFP) sponsored by the National Science Foundation (NSF) is arguably the most popular fellowship for STEM PhDs. I was fortunate to win a GRFP, but the process was by no means easy or transparent, so I thought I’d jot some thoughts down to help guide future students. All views are my own.

Why a fellowship?

Most important takeaway: The selection process for fellowships is incredibly random and whether or not you’re able to secure your own funding has no bearing on your future success in the PhD. ✨

OK, now that I’ve said that, I’d like to first point out the advantages of having your own funding over being a Research Assistant (RA) on your advisor’s grants or being a Teaching Assistant (TA).

  • Flexibility: As an RA, you must contribute to the research topic that was proposed in the grant application (which you likely did not write) or else your advisor gets in big trouble. If you like that topic, great! But maybe you don’t and would prefer to work on something else, even if you don’t know exactly what “something else” is (like me). Whereas grant money goes towards projects, fellowship money goes to individuals, which means you can research whatever you want, even if it differs from what you wrote in your application materials. The NSF knows that research directions/interests can rapidly change and they only care that you’re being productive with their funds, not as to what you’re researching.

  • Peace of mind: Most of the large fellowships like the GRFP are multi-year fellowships that can fund the majority, if not all, of your PhD. Therefore, you can focus on your research without stressing about applying for other grants or excessively TAing. If you like teaching (like me) you can still TA, but it’s just not required.

  • Comfort living: While you won’t get rich off the stipend, the GRFP does pay a reasonable salary (currently $34,000 a year) that is higher than the stipend at most schools. If your program pays more than this, ask your department administrators as to how they can make up the difference, as most schools should be able to!

  • Resume boost: All large fellowships are well-established and fairly competitive, so listing it on your resume will greatly elevate your status in the eyes of future employers in both academia and industry.

  • Expand your network: Some fellowships (not really the GRFP as it’s too big) have a strong community and the funding agency hosts annual conferences where you get to meet other fellowship recipients. This gives you a chance to learn about what other cool research is out there and share your work with others!

  • Cultivate skills: The essays (more on the NSF essays below) are tedious but they give you a taste of the type of writing you will have to do during and after the PhD. There’s usually some form of a personal statement where you have to describe your research background and a research statement where you will have to write a new project proposal with motivations, methods, and impact.

  • Prepares you for PhD applications: If you’re not yet in a PhD program, then these fellowship applications are good practice for the PhD applications, as the fellowship apps are typically due earlier. For example, I copy+pasted my GRFP essays verbatim for most schools.

  • Can affect admissions: If you are waitlisted or rejected by a PhD program, but then later find out that you’ve won a fellowship, then several programs will then change their decision and admit you because you’re now able to fund your own PhD. You shouldn’t bank on this happening, however, and I strongly believe you shouldn’t choose a program that didn’t [initially] choose you. But this is a personal preference.

  • Your advisor will tell you to: Advisors want to save money and help you build your resume, so they’re going to strongly suggest (i.e. force) you to apply. If you take the initiative, then that’s brownie points. Even if your advisor has research funding guaranteed for you, it doesn’t hurt to still apply and potentially gain even more options!

Logistics

Here’s a bulleted list of logistical pointers about the GRFP that I’ve distilled from information that you can also find elsewhere. Please double check that this information is still current.

Deadlines(‼️) ⏰

The GRFP is typically due the last full week of October, with the date differing by discipline. Materials Research is typically due on the Tuesday of the week. All materials are due at 5:00PM local time and they are very strict about this! Make sure you and your letter writers know. They do not make exceptions and you don’t want to be caught on the wrong side.

Eligibility

  • Class year: Essentially, you can apply once as an undergraduate student and once as a graduate student. If you don’t apply as an undergrad, then you don’t get that opportunity back, so it is in your best interest to apply as an undergrad if you’re going directly into a PhD. If you have a joint BS/MS degree (like me), then their new policy is you can only apply once as a “graduate student,” unfortunately.
  • Topic: It has to be STEM-related, but even then it can be pretty flexible (e.g. basic science, engineering applications, STEM education, possibly even STEM policy-related?). However, it seems like starting in 2020 (NSF 20-587) there are new high priority research areas?? I have no idea how these will be prioritized.
  • Citizenship: If you’re at least a permanent resident then you’re good.

Statistics

There are over 12,000 applicants and about 2,000 GRFP awards, so you have a 16% chance overall. These are actually pretty good chances compared to other fellowships. However, they also strive for “diverse representation” in terms of applicant background, geographical area, institution, and Field of Study. This is both a blessing and a curse.

For example, my field was “Materials Research - Materials, other - Computational Materials Science” and I’m pretty sure they will only fund one of these each year, no matter how many great applications they receive. Furthemore, it’s well known that these fellowships disproportionately go to students from the top schools, so that’s another thing to keep in mind, but don’t let it get to you. Basically, just be the best and none of this matters. 😁

Materials

Pretty standard. Demographic information, transcripts, three letters of recommendation (they say two will be considered but you won’t win unless you have three), and two essays.

Letters of recommendation

Again, you must have three if you want to win, and they must all be submitted by 5:00PM at the deadline. No exceptions. You can ask up to 5 writers if you feel like one might drop the ball, but extra letters beyond 3 won’t be considered. Some things to consider:

  • When to ask? At least one month in advance. Follow up with biweekly reminders until you see in FastLane that they’ve submitted it.
  • How to ask? A simple email that bolds the deadline and both review criteria (maybe even with descriptions/definitions). It is also extremely helpful to have essay drafts for them to skim so they know how to best support your statements. Good writers will care enough to do this. Mine did. 🥰
  • Who to ask? Ah, the tough question. You must have one from your current advisor (undergrad or PhD), even if you just met them as a first-year PhD student. Otherwise, reviewers will be confused. The other writers must be able to speak [ideally glowingly] to your research potential in terms of both review criteria. They don’t have to be professors or even have a PhD. I asked my work manager (BS EE, MA Education) for a letter because I felt like he would offer a perspective not found elsewhere, particularly for Broader Impacts.

List of winners

You can see and search through a list of past winners on the GRFP website, unless they requested anonymity. Most people show their names—it’s a great honor! This list can be helpful for you to find past winners in your school/program to ask for tips if you seek help but don’t know who to ask.

Essays 📝

To help contexualize this section, here are my essays: personal statement and research statement. Caveat emptor applies.

Deep inhale. OK, here we go. You’ve finished all the boilerplate and it’s time to sit down and write those essays. There’s a lot of suggestions out there if you do a quick Google search, so to be as concise and non-redundant as possible I will just make a bulleted list of rapid-fire points for what I think makes winning essays.

General pointers ☝️

  • Start early so you have time to proofread and have a draft for your letter writers.
  • Get friends/peers who have won the GRFP to proofread your essays. Ask them for targeted feedback.
  • You must follow the guidelines. If they say font type, font size, margin size, etc. you must follow it precisely. If you’re lucky, the system might tell you there’s an error you have to fix. If you’re unlucky, the system will reject your submission outright without review and without the option to resubmit. They are very strict about this. Follow the guidelines. Just do it.
  • Both of your essays must address both review criteria, Intellectual Merit (IM) and Broader Impacts (BI). They require separate headings for each one. Again, just do it.
  • Bold important phrases and sentences to make them stand out. You’ll be writing paragraphs, but reviewers will want to know what’s important. I was definitely not used to doing this and thought it looked jarring, but it is what it is. This will greatly improve your chance of winning.
  • Include figures, even if it’s just a methods figure. Results would be even better. I had two results figures in my personal statement and a very straightforward methods figure in my research statement. I’ve heard “no figure = no award.”
  • Include references in both statements to build ethos. The formatting is up to you and you can get away with space-saving measures here.
  • Focus on your contributions and talk in first person as much as you can. Walk the reviewer through your thought process.
  • Use numbers and highlight impact, not just intent. This shows your contributions to society in a way that’s easily understood.
  • Be honest (duh) and write candidly. Use precise language and directly state what you did without too much embellishment or flowery language. The reviewers are scientists, not undergraduate admissions officers. Define all acronyms.

Personal statement

  • To my knowledge, most people write chronologically. While this isn’t a requirement, it probably makes the most sense as your educational trajectory probably tracks your research impact.
  • You want IM and BI to be pretty much a 50/50 split. When they say both criteria are important, they mean it.
  • For IM: focus on the research. Don’t discuss extracurriculars and other non-STEM stuff too much. Or at the very least highlight methods, skills, and results that could translate to the STEM research domain. Again, focus on your contributions, thought process, and impact.
  • For BI: Typically this means outreach, particularly for initiatives that increase diversity in STEM. These don’t have to be directly tied to your research, but they should ideally have some ties to STEM. Try to work in as much as you can and BI should be at least one full page.
  • You probably don’t need to discuss anything from high school, unless you’re applying as an undergrad and you have a very formative experience (e.g. published paper, international olympiad medal, ISEF/STS, personal background) that ties into the story arc. Keep it short though.

Research statement

  • Feel free to use a lot of headings to guide the reader, much like a journal paper.
  • This should be about as technical as the “Introduction” section of a research paper. Your reviewers aren’t necessarily experts on your topic and an important part of the scientific discourse is being able to appeal to scientists of all backgrounds and disciplines.
  • Unlike the personal statement, the BI section here can be a lot shorter (just 3-5 sentences). It’s also less about personal outreach and more about how your actual research might address things like energy/resource conservation, socioeconomic inequality, the future workforce, national security, climate change, public health, etc.
  • Have your advisor proofread it as they’ll be more familiar with what language appeals to the NSF. Mark definitely gave me a lot of tiny but helpful pointers. They also have to address in their rec letter how “original” your statement is and how much help they gave.
  • One question people ask is, “Should I ask my advisor to brainstorm a good research proposal?” My answer would be No. For one, there’s the originality point brought up before. But even more so, I challenge you (as I have challenged myself) to come up with a good proposal and do your own research for what to include. You’ll have to do this in the PhD. You’ll also feel more passionate writing about a plan that you created yourself and this passion (or the lack thereof) will be readily apparent to a keen reviewer. You should get your advisor’s suggestions only after you have a draft assembled.

Reviewer comments

The cool [and intimidating, and gratifying, and demoralizing] part about the GRFP is that you get to see reviewer comments on your application, mostly focused on the essays. There are three reviewers and a column for each review criteria, so BI really is as important as IM! Peer review is a big component of the research endeavor, so I thought it was great to get this experience, even if I didn’t get to offer a rebuttal. Furthermore, I noticed something interesting…

So I applied twice for the GRFP, back when joint BS/MS students were still allowed to. The first year I got Honorable Mention and the second year I won. In my first set of reviews, they were terse in their assessment but all positive (at least “Very Good” and no negative comments), so I was a little surprised I didn’t win. In my second set of reviews, they were a lot longer. I almost died (in the bad sense) when I read the first review because it was quite critical and demeaning. Then I almost died again (in the good sense?) when I read the last review because it was incredibly glowing and effusive. And so my conclusion is that in order to win, you need to get one reviewer to really back you up, and the others don’t matter as much? Another friend of mine had a similar experience. Of course, there’s no way to predict their reactions ahead of time…

Final thoughts 💭

Wow, if you actually read this whole thing, h/t to you. 👏 I hope this helped. There’s also a lot of great advice and sample essays on Alex Lang’s website and if you want to get stressed out while waiting for results (or commiserate with others…), check out The GradCafe forums.

~ Good luck!!! ~