NSF Grants
sorm...@member.ams.org
proposals came to CUNY yesterday. Basically the company consists of
retired men from the NSF, NSA and other government agencies and they
provide pointed advise about how panels view proposals and help
faculty rewrite theirs in a way which will better sell their ideas.
The same company is available for hire at other universities or even
by individual faculty.
According to the representative, who is by the way, a math/CS person,
the key is to take the proposal of someone with strong credentials and
serious content and make it watertight. He said the panel is a group
of mathematicians trying to look smart; their favorite thing is to
find some bug in a proposal and drop it at the bottom of the pile.
Other important points he made:
1) Always format the proposal exactly according to the program
specifications with clear titles for each section: "Prior support"
"Broader Impact" and don't use clever titles for these subsections:
make it easy to find them.
2) Make your proposal have a real focus that can be described in a few
sentences and put them in the summary. Many panelists only see this
before they vote based on the recommendations of the three. Thus the
summary should be written a bit more for experts than officially
recommended by the NSF. The title should match this summary well.
3) Feel free to quote liberally from the program description and then
describe how you are going to address the point, but focus on
deliverables not just propoganda.
4) Have a critical colleague look over an early draft as well as the
final proposal. Some universities have grants departments that help
with this, or you could ask your university to hire a consultant (eg
their company) to do this. Having someone go over a rejected proposal
with you can help as well and not just the Program Officiers in your
field but someone who can give you serious honest criticism of the
selling value of the proposal.
He also recommended a book by Paul G. Chapin "A guide for scientists
seeking funding".
He strongly emphasized that all this advise is targetted towards
people who are doing excellent fundable research and just have not
been getting grants. It isn't going to turn a proposal without
"content" into a fundable proposal. But he did claim that his company
in particular had helped improve the grants success rate in a variety
of states, the so-called EPSCOR states. His company has EPSCOR
funding to assist faculty in these states.
sorm...@member.ams.org
at a less experiences group:
http://www.aimath.org/mathcommunity/funding.html
Key points quoted from there:
-----------------------------------------------------------------------------------
Content of Your Proposal
* Mathematical: In your project description, propose a range of
research problems that you intend to work on. You can be speculative
on some, especially in the latter sections. Be clear about why a
problem is interesting, and how you propose to solve it. Give
sufficient detail to convince the reader that your project is feasible
and will lead to success. Propose some problems that fall within the
scope of research you have already done, to show that you have the
necessary expertise/experience. Also include more ambitious problems,
explaining their significance and why they move your research area
forward. Explaining how you may possibly ``hit a home run'' with a
potentially huge conjecture will draw attention to your proposal,
allowing it to stand out from the rest.
* Budget: You should think carefully about your budget, and you
will need to coordinate this with the grants officer in your Sponsored
Research Office. You may need to argue to reduce budget amounts in
certain categories, if you feel they are not appropriate for your
individual proposal. Apply for what you can use and can justify. In
general, proposals are reviewed and ranked based on their intellectual
merits, not on proposed budgets. Funding is awarded based on rankings,
and if your budget is too high, the funding agency may simply scale
back your request. Don't apply for too little, out of embarrassment.
The support your institution receives from external funding is often a
key to important things such as keeping your school's teaching load
low, or providing conference travel funding for post-docs.
* Broader Impact: An important component to our profession is work
which enriches the mathematics community, at a variety of levels. You
may be involved with an REU, or with individual research with
undergraduates or graduate students. You may be involved with outreach
to talented high school or middle school students. In discussing the
``broader impact'' of your proposal, describe both the things you have
already done, and your future plans for getting others involved with
mathematics. For NSF grants in particular, discussion of broader
impact is a nontrivial part of your proposal and must be addressed.
However, your outreach need not correspond closely with your own
research problems. For example, you may be working with students on
other problems or on expository research as part of your department's
honors thesis program. Providing a positive, research-like experience
for students is valuable, even when it doesn't lead to original
research. Describe your involvement, and request appropriate funding,
such as support for undergraduates or graduate students to attend a
conference. There are many other reasonable activities you could
suggest, which get students involved with and excited about
mathematics! For example, if you are bringing in speakers to give
talks to the Math Club, you many want to ask for funding to support
this activity.
* Biographical: Your biographical sketch is a good place to
describe the broader impact of your proposal. For NSF proposals, this
falls under the heading of ``Synergistic Activities.'' You've done
these things even if you don't realize it! For instance, you may have
organized undergraduate colloquiua in your department, or served on a
student's thesis committee. Ask a senior colleague to help with this
section, to know what experiences to highlight.
Maximizing Your Chances for a Successful Proposal
* Know your audience: Grant applications may be reviewed by panels
or individual reviewers, by experts in your field or a general
mathematical audience. Find out as much as possible about the way in
which proposals will be reviewed, and tailor your proposal as
appropriate.
* Get feedback from trusted senior people: Definitely ask others
for advice, and don't be shy about it! Consult senior people whose
opinions you trust, and have them read your proposal. You need
substantial feedback, not just ``Oh yeah, it looks great.'' You should
seek feedback from several people, including (if possible) a colleague
who has been a successful grant applicant or who has reviewed grants
in the past.
* Use past reviews to improve your proposal: Sometimes a rejection
letter comes with almost no feedback, but it is more often the case
that an unsuccessful proposal will have specific feedback from
reviewers. These can be useful, although be aware that individual
reviewers' comments may sometimes be contradictory. Read your reviews
and incorporate useful comments as you write your next proposal.
Comments regarding level of detail, style, broader impact, etc. are
often valuable and will allow you to craft a more competitive proposal
the next time.
* Establish credibility by describing your prior research: If this
is your first grant, it is important to establish your research
credibility with reviewers. Use the first few pages of your proposal
to describe research and outreach that you've already done. (For
instance, use 5 pages out of the 15 page limit for your Project
Description, in an NSF application.)
* Demonstrate breadth, knowledge of the literature, and an
expanding research program: It is important to show the reviewers that
your research program is growing, and expanding into new areas.
Convince the reviewers of the breadth of your knowledge in your
research area. In particular, show that you're aware of contributions
of other researchers in your field. Nothing puts off a reviewer more
than when his or her work is relevant to a proposal but it hasn't been
mentioned!
* Go to conferences and be visible: Enhance your visibility to
potential panelists and reviewers. Get out there and be known! If a
reviewer has seen you give a research talk at a conference, for
instance, this puts a face on your proposal and may help it stand out.
Also, have a website with your preprints available. Mention in your
proposal that your papers are available, and give a link to your
website.
* Be persistent: Keep applying! Learn from past grant-writing
experience, and be immune to criticism and you'll eventually be
successful!
-------------------------------------------------------------------------------------------------------
sorm...@member.ams.org
The following points were crucial as not addressing them could lead to
the dropping of a proposal:
Broader Impact: cannot be to the sciences, must be to the community at
large, education of your students, mentoring women, math club talks,
and must include ideas for the current project not just past
performance.
Applications of your research to science should be mentioned within
the proposal and are part of the Intellectual Merit not the broader
impact.
Be sure to document collaborations.
Make it exciting up front so the reviewer isn't too bored to read it.
Play up your strengths, explaining how your past research justifies
your ability to proceed with the particular plan.
Use your institution's strengths: really address how you are better
able to proceed with your project because of XYZ about your
institution. If your institution is not top ten, you may wish to
describe the place in more detail, and who works there and
communicates with you.
Proposals that ask for funding for graduate students are viewed
favorably. Document the graduate students. Mention them by name and
list past doctoral students.
Have someone critique your proposal, preferably someone who has been
on a panel in your field. People with prior NSF support are likely to
have served on an NSF panel or two. Panelists are not allowed to
mention when they served on a panel but they are allowed to say they
have served on a panel in the past.
Call your NSF program officers (rather than emailing them) to ask
about the panels: what subfields will be on the same panel as your
proposal? Can a certain activity be counted as broader impact?
Anything.