The NSF CAREER award is perhaps the most prestigious young researchers award given by the National Science Foundation (NSF). The PI of a submitted proposal must be on the tenure track, but untenured. The award is over a five year period, and the minimum proposal budget is $400,000. There is no upper bound, but on average the amount is around $500,000. A PI has only three tries to get funded. Finally, the proposal itself must consist of a research proposal and an integration of research into education proposal. For more of the details please see the NSF solicitation.
After all, one can only get this award once, and only after three tries. Most likely everyone has an idea of what they want to write a proposal on, but framing it as a CAREER award can be difficult. As is finding the balance between how the research proposal impacts the education proposal and vice versa. Another question is how “grand” should a proposal be? The answers to these types of questions can really help one write as competitive proposal as possible. The workshop at NSF is geared towards answering these types of questions.
We were told that there were 515 applications with enough seating for 200, but they increased capacity to 300.
The remainder of this post consists of notes on each agenda item for the day. I am going to try and record what I found to be most useful along with some of my own comments. All of this is in my own words, and any inaccuracies fall on me alone.
Getting into the building was not as bad as the organizers made it out to be, but I got here right at 7am; the time they recommended. It is exactly like airport security as long as one has either a drivers license or a passport. The security guards I encountered were polite and nice. They gave each of us badges with our names on it, and we headed up into a huge room on the second floor.
Attendees were required to know the basics of the CAREER proposal prior to arriving to the workshop, and so the organizers started right in with the details.
One question that program directors hear a lot is, when is the best time to submit a CAREER proposal? The answer is simple, when you have a good research idea and a well-developed, well-written proposal with an important problem, a compelling means of solving it, and promising preliminary results. If you do not have these, then wait, or else submit. It does not matter if it is your first year as an assistant professor or your fourth.
Some surprising statistics were given:
The NSF is pushing CS education forward by launching a new program called CS+X. This new program seeks proposals for teaching CS concepts in relation to other fields. Not just programming. For example, one might be teaching computational concepts to chemists. Those interested should be on the look out.
If you do not already know there is a concept called a Dear Colleague Letter at the NSF that communicates new programs, etc. Furthermore, sign up for email alerts.
This session started with some more overview, but then went into a Q&A. Here is what I learned:
On the one page summary make sure to add keywords. These are actually very important. Add two to three keywords that describe your project. These are used for routing between programs and finding reviewers.
The program directors are very committed to insure that each proposal is placed at the appropriate program. If you are unsure which to submit to, then contact via email the various program directors at each of the programs you are interested in, and they will help you figure out where to go. If you are still unsure, then pick one, and they will internally route your proposal. Some proposals may need to be evaluated by more than one program, and this is okay, but must be specified on Fastlane.
Now don’t spam program directors: after sending an email wait 48 hours for reply. If you need to contact more than one program director, then CC everyone at once; this makes sure everyone knows of the communication which can prevent needless internal communication. When emailing program directors include a one page summary of your project, and be clear about the science contributions in the one page summary.
When writing a proposal understand the audience. Make sure you know who the scientific community is you are writing to and then word the proposal for the program you are submitting to.
As I said above, the budget is a minimum of $400,000, and this usually pays for about one month summary salary for the PI and one full time PhD student. This is the very bare minimum a proposal should ask for. However, don’t feel obligated to be at the minimum. Ask for what you need to be successful.
Collaborators are allowed to submit a letter of collaboration as part of a CAREER proposal; however, be very clear that this is the PI’s CAREER proposal, and make it very clear, that the collaborator is there to support you, but this your idea and your vision. Clearly specify the role of collaborators in the proposal.
On the relationship between CRII and CAREER. These are two very different awards, and do not have to be linked. Furthermore, a CAREER PI does not have to have the former to apply. The former also does not increase your potential for funding as a PI. The CRII is a two year award useful for getting initial results, but the CAREER is a five year award that is meant to launch ones career as a researcher.
A lot of people make the mistake that they write a proposal that essentially just carries on with their dissertation work, but the CAREER was designed for new PI’s to get out from underneath their advisors. Your proposal should envision your future as a researcher. This is the time to switch areas if you want to switch areas. This is the time to use your jewel of an idea. This is the time to dream. Don’t just do your dissertation, do you!
Keep in mind that you can only submit one proposal to only one primary directorate, but many programs.
If you were declined:
DO NOT BEAT YOURSELF UP! I know first hand about this, because my first round was declined, and I felt like a failure; like an imposter. But, going to this workshop made me feel a lot better, because a fast majority of the people I met were also declined. Nearly every panelist got rejected as well. You are not alone! Don’t give up, and never stop trying!
When you get your reviews take them in, then leave them alone and come back to them. Then go over them again. Once you understand them you will be tempted to just want to fix up your proposal like a paper, but resist this. Time and time again we heard this at the workshop. Rewrite the entire proposal. Even choose a different title. Rewrite, rewrite, and oh, did I mention this? Rewrite!
Do not assume the same panelist will review your proposal again. It is more likely that the panel will be different than the same.
The review panels selected will be broad, but program specific. So your proposal should fit that audience. Don’t assume they are experts in your field.
In your proposal you should include a related work section. Make sure and be very clear about what’s been done in the literature, or the reviewers will fill in those gaps for you, and it may not be good for your proposal. Don’t make your reviewers infer what you mean; they cannot read minds. The proposal should be complete, to the point, and fluid.
Timeline is important. Try and create a vision of what you are going to do. Do not get super specific. For example, do not break it down into months or years, because you might not get something done that the rest of the proposal depends on. Use phases or thrusts instead. You cannot predict the future, but you should be able to give a convincing timeline of your work. Think five years or even ten years ahead. As a researcher we should be doing this anyway.
Each phase of your proposals timeline should have an evaluation plan, and within these evaluation plans you should discuss alternate paths to success; have a plan A and a plan B when plan A fails. You want the reviewers to know you will be always making progress.
It can be tempting to spend a lot of space in your proposal talking about preliminary work you have done, but resist this, instead spend most of your time on the new stuff! Limit your self to just two pages about preliminary work.
Again, be clear about what’s new. The CAREER is the jewel!
On education:
The proposal should have a clear integration of research and education. Integration is so important. Education: Creditable plan, don’t do a boilerplate!
One good piece of advice we got was use existing resources at your institution in your education plan to make it impactful. Link up with your universities or colleges existing initiatives to help your proposal impact the area you are trying to impact. For example, get the college of education on board to help evaluate new curriculum, find K-12 resources at your school, do the leg work.
Do not put this off to last minute.
SHF program directors said that a borderline proposal may be funded due to the education plan, but a bad research proposal with an outstanding education plan will not get funded. So it is in our best interest to do a good job on this.
Proposals should advance the state of the art, and make sure you convince the reviewers that you do know the state of the art, do not just cite a bunch of papers, actually read them.
Try and answer the question: Why are you going to spend the next five years doing what you are proposing? Then write about it.
Be passionate!
Proposal should be novel, and not connected to existing or prior funds, but it can build on prior work. We were told time and time again that you do not want even a hint of overlap with prior work. Make sure this is clear in your proposal.
Again, this is a career launching project: vision is important.
At the end of your proposal have a future vision paragraph that talks about where you are going next when this proposal is done.
This is what I am trying to answer right now: what is my long term vision? You should to.
In the next two sessions we heard from various winners over the last couple of years. These were interesting, because they had a lot of overlap; which conveys what is important and common between all winners. Here is what I learned:
In this first session every speaker submitted more than once. One person three times.
This is a big one: start early. Six weeks is not enough! Give yourself at least three months, maybe even more, because you might need to get letters of collaboration from government, schools, or industry. They all take time. You will also want to get feedback on your proposal. Never submit any proposal that you were the only one that read it.
Make time to polish the entire proposal.
Make a plan regardless of funding. This should be a project you are passionate about. If you don’t get funding, then having a plan of action will allow you to get preliminary results prior to the next round, and work on rewriting.
Make it fun.
Get feedback from people:
Keep the writing process structured. Plan to write. Know when and where. Stay motivated.
Keep everything in mind and not just the project description.
Treat education just like research: doesn’t have to be a big new thing, but make it interesting by connecting it to research.
Proposal should have focus. Why are you doing this? Why should it be funded now?
This was brought up a ton!!!! Sit on an NSF panel to better understand the other side. New researchers can sit on NSF small and medium panels. Contact your program director about this.
Proposal should not be a laundry list of things to do, but should be a story. Push the vision to the front.
To get feedback, give talks on the topic at conferences.
We heard this a lot too. The first few pages – two to four – should be exciting! Make the panel reviewer jealous. Make the reviewer want to share your work.
This was brought up, and is nice to hear, we should identify weaknesses in the proposal. Don’t let the reviewers find them.
Get as much feedback as you can.
The section titles should convey what you are doing.
This session had past winners who won their award on the first go. There was not many new insights here. Most of what was said in the previous session was reiterated. So go back and read the above again. That tells you how important this stuff is!
One thing I liked that one speaker said was, every page you write is like $50,000. Make it count!
Use metrics to evaluate your education plan, and make use of experts that can help you do this. Just takes some emailing the right departments at your institution.
Have milestones, evaluation methods, and think about sustainability
Big picture, well structured plan on how they will achieve it, then technical details.
Be concrete using examples.
The career awardees then opened it up for Q&Q. Here is what I learned:
Another one that was repeated a lot: talk to your program directors.
Use graphics to explain your main idea. Explain a range of approaches to solving your problem not just one linear path, or else, the reviewers will not understand how you will get out of a problem.
I was really glad to hear this: don’t worry about who you are or where you are.
Might be a risk to being very specific in your proposal. Keep it high level.
Make sure broader impacts section is clearly marked; this is part of the PAPPG.
Be repetitive: hammer your ideas into their heads.
In this session the program directors gave presentations on what they felt was most important. Here is what I learned:
Don’t waste tries you only get three.
Proposal are rated as follows:
This was said a lot, the NSF doesn’t care about incremental research. This should be visionary, but doable in five years.
If your school does not have any PhD students, like mine, then a postdoc should be highly considered.
Proposal should flow! Get the reviewers to want to pass your proposal on.
Show them you have a plan, not what you are going to do.
The Top Ten Mistakes (starting with the least severe):
Too small of font.
Figures are illegible
Acronyms and Abbreviations
Dissing the competition
Poor distinction between preliminary results and proposed work.
Clearly mark preliminary vs future work so that the reviewers don’t get confused.
What is your end goal? What stands between you and the end goal? What do I bring to the table to convience the reader that I can over come the obsticles.
Don’t use buzz words if they are not what you are working on.
Lackluster education plan
“It wasn’t clear…”
Confining yourself to your PhD work
Research Plan lacking Cohesion
How to serve on an NSF panel:
Program Directors hate it when titles don’t reflect the content of the proposal.
Finally, we had program specific sessions with the program directors. I went to the SHF session. Here is what I learned:
Leave a comment by sending me an Email