The National Science Foundation Graduate Research Fellowship (NSF GRFP) application season is here. I was fortunate enough to receive a fellowship for my application is 2017 after an unsuccessful one in 2016. From one rejected and one accepted, here is a list of resources that helped get me through it, a bit of my personal experience, and tips for applying including reviewer comments.

What is it? Should I apply?

The NSF GRFP gives STEM grad students full tuition and stipend for three years as well as a few other opportunities and resources. If you are an undergraduate looking into graduate school or an early graduate student, you should definitely think about applying, although make sure you are eligible (eligibility guidelines). If you are a graduate student, you can only apply once. There is not necessarily a right answer as to whether you should apply in your first or second year; the best thing you can do is understand the application and then decide whether or not the timing is right. Keep in mind that you are judged against those in the same status (i.e. first-year grad students are only compared to first-year grad students) so if your application will not be substantially stronger your second year, the first year may be a good time to apply.

How to Apply

Every NSF grant including the GRFP has a program solicitation (link). Read the entire thing and then read it again. It has almost all of the information you need on how to apply. You should essentially commit the application instructions and review criteria sections to memory. To summarize though, you will need a personal statement, research statement, and three letters of reference. You are evaluated based on two criteria: intellectual merit and broader impacts. Intellectual merit relates to the proposal’s potential to advance knowledge whereas broader impacts relates to the proposal’s potential to benefit society. These are the only two criteria and they are scored equally so they both must addressed.

Most applications struggle with the broader impacts section, and it is common to be confused about what broader impacts actually means. Fortunately, the NSF put out this letter detailing what kinds of activities address broader impacts. It is definitely not a bad idea to use the same terminology bolded in that letter for your broader impacts section (see below about subheaders).

Also read everything on the GRFP website (link) under “Applicants” and “Reference Writers”. You should also read this several times. They have extra information here on how you should write your statements as well as tips for applying. Under reference writers, the requirements and tips for those writing your letters of reference are detailed. There is a good chance that your reference writers will not read any of this so make sure that the important information there is conveyed to them.

Still have questions after reading all of that? Fortunately, the NSF has a FAQ section for this program here.

Non-NSF Resources

• Alex Lang’s website (link) – This is probably the single best resource not from the NSF. The advice and tips are all very good. Perhaps most useful is his collection of successful essays. Keep in mind that the format changed in 2014. Also note that just because an essay was awarded, it does not mean that yours needs to be just like it. If you see problems with an essay, especially if it does not follow the solicitation guidelines, you should not do the same. For example, several awarded essays combine the “Intellectual Merit” and “Broader Impacts” sections in the Research Statement even though the solicitation asks for separate statements. I tried this approach my first time around and one reviewer really did not like it. Again when in doubt, follow the solicitation guidelines.
• GRFP Essay Insights from Robin G. Walker, University of Missouri, Columbia (link) – More application information and writing tips, most of which are very good. If you find outlines and self-assessment rubrics helpful, she provides some here.
• reddit.com/r/gradschool – Tons of threads here with personal stories about people’s own GRFP experiences (just search ‘GRFP’). This subreddit can be useful to quickly read what other people went through and what kind of reviewer comments they got. However, try to avoid getting sucked into the general depressive attitude.

My Personal Experience and Thoughts

• The Research Statement – Be focused to provide enough detail. Propose something feasible with clear end goals, questions, and hypotheses. If you can, explain how you are uniquely suited to accomplish this research. In my first year, my research statement could have been a lot better. As the reviewer’s noted:
  • “The proposed work [is] too exploratory and the hypotheses [are] diffuse”
  • “I was not convinced that [these] experiments…can be useful.”
  • “Structure of the research plan needed improvement.”
  • “The research plan would have been more coherent if each objective had been dealt with in its entirety.”
  • “The application relied too heavily on a small number of citations to support multiple statements.”
  • Presenting the intellectual merits and broader impacts…in a isngle section resulted in the significance…being somewhat lost.”
  • “Providing opportunities to students with limited access to STEM disciplines could have been improved.”
  • “The letters of reference could have been improved in terms of making a case for how the application could specifically benefit from the fellowship.”

Again, it needs to be clear and concise. Make it easy on the reviewers to review it. The biggest change in the second year was just restructuring and brining more focus to the proposed research. This is how I laid it out each with a bolded header:

• • • Introduction – 3/4 of a page ending in a bolded one-sentence statement describing the research objective.
    • Proposed research – 1 paragraph again explaining in one sentence what the goal is, one sentence for how you’re going to accomplish it, and then a list of research aims in bold and numbered. I had three aims explicitely labeled in the sentence, “I am to (Aim 1) …. (Aim 2) … etc.”
    • Hypotheses – Clear hypotheses corresponding to each aim in a list, e.g. (Aim 1) Hypothesis for Aim 1. (Aim 2) Hypothesis for Aim 2. etc.”
    • Methods – Two paragraphs describing your approach in fairly easy to understand terms and from a high level. Make sure you describe that you have the resources and skills available to conduct the research.
    • Intellectual Merit – Why is this scientifically relevant?
    • Broader Impacts – What activities related to this project will benefit society. I used itacilicized subheaders to make this clear within the borader impacts categories listed in the letter above, e.g. I wrote, “Broadening participiation: …. Promoting education and scientific understanding ….“
    • References – Use a short format here to save you space and allow you to supply enough references to support your statements. In-text, I used superscript numbers. In the reference list here at the bottom, here is an example, “[1] Author 1, et al. (year) Abbr. journal name Vol:pp-pp. [2] Author 1 & Author 2 (year) etc.”

Even though the reviewers will be from your field, it does not mean they will understand all of the jargon you are used to hearing. Definitely get at least another scientist from outside your field to review your application and provide feedback. They should still have a good understanding of what you are talking about.

Here are some of the positive comments I got. Try to write something that will lead the reviewers to the same conclusions:

• • “The proposed research is timely and addresses interesting questions.”
  • “His background…would enable him to conduct the proposed experiments.”
  • The proposed research plan…has a clear research question, and testable and predictive working hypotheses.”

• The Personal Statement – Everyone’s background is different so it is hard to give advice and say what to do here. Panelists also have wildly different opinions on what is good and what is bad here. I’ve had one former panelist tell me how much they love the “ever since I was a child” essays and another tell me how much they despise those. Again, definitely follow what is in the solicitation. Explain what your career goals are. You do not have to want to stay in academia to get a GRFP. One of the main goals of the program is to create an educated STEM workforce here in the US. There is a lot of misinformation out there about this statement which is why it is so important to talk to experienced panelists, again understand the solicitation, and filter out statements that contradict it. Here is how I structured it (headers in bold):
  • Introduction – Two paragraphs about how I ended up where I am and what my end career goals are.
  • Intellectual Merit (Relevant Background) – About one page on my past research experiences and results of the described research. Explain what specifically you contributed and how you were valuable if you collaborated on something.
  • Broader Impacts (Relevant Background) – About one page on my past broader impacts activities again using subheaders like “Advancing Discovery and Understanding” and “Broad Dissemination to Enhance STEM Understanding.”
  • Future Goals (Intellectual Merit and Broader Impact) – What are you going to accomplish with this fellowship and beyond. How you are uniquely positioned to be successful.
  • References – You may not need to include these if it doesn’t apply but I did using the same format as my research statement.

Even though I was awarded my second year, I still got two “Very Good” scores for Broader Impacts. One reviewer noted that “More detailed discussion of the leadership roles in outreach activities will help distinguish this aspect from other applications.” Similarly another noted, “The application could improve…by detailing how his involvement uniquely enhances the programs he’s working with.” In other words, it isn’t quite enough to just describe that you were involved with activities X, Y, and Z. You should include how specifically you contributed to these activities and how your involvement was valuable.

The GRFP is extremely competitive. The NSF can only award 2,000, and the chance of there being more than 2,000 excellent applications is high. The sad reality of the program is that many excellent applications are rejected or get an honorable mention. Bad luck with one bad reviewer can make the difference. My second year’s application was really boosted by having a submitted first author publication (all three reviewers made this a major point). In both year’s comments, it is apparent that reviewers use publication as an easy way to separate people out which is unfortunate for many that have not had the best opportunities. If you can, try to have something at least co-authored and submitted by the application deadline. At the end of the day, luck has a lot to do with it which I know is frustrating to hear but is unfortunately true. Follow the solicitation, try your best to write the best application you can, and keep your fingers crossed. Good luck!