Transcript Document

“Idea to Grant”
Beginning the search for research funding
John Eiler
Geological and Planetary Sciences
March 16th, 2004
Outline
• Should I write a proposal?
• When should I start, and how long will it take?
• The care and feeding of program officers
• Brainstorming, caucusing, and pilot studies
• Outlines
• How much money do I ask for?
• Seven deadly sins; five golden rules
Should I write an NSF proposal?
Do I really need money?
Do I have any other grants or ‘funny money’ that can be used?
Is it worth the wait and trouble?
Does the project connect to the goals of an NSF program?
Would it fit better in another agency?
Is the project mature enough to compete effectively?
Proposal-writing is about trading time for money.
Both are valuable; always be sure you trade to your advantage.
Should I write a renewal NSF proposal?
Is this still a good problem ?
Did I fail, or just not quite finish?
Has this morphed into a new project?
Will my group stagnate if I keep working in this area?
Does my program do ‘accomplishment based’ renewals?
Picking through the trash
Re-animating rejected proposals
• A large fraction of rejected proposals are not good and should die a quiet death.
• Re-submissions are supposed to wait 1 year, and should involve substantial
revisions and improvements.
• Always respond positively to
criticism.
• Reviewers and panels have short
memories.
Wait for Tony’s talk for more
specific advice!
“Look, mom, last-year’s failed MRI proposals!”
When do I start ?
• There is a ca.6 to 8 month lag between when you sit
down to write and when money shows up at your door.
• You can sometimes spend money ca. 1-3 months
before it arrives on campus. Don’t count on this, and
don’t make a habit of it.
• Pick a panel and deadline ca. 2-6 months before
submission. Six is good; two is common but usually
produces sub-par proposals.
• The most important activity you can pursue in the
months before writing is the ‘pilot study’—a solid,
first pass at the problem.
Brainstorming, caucusing and pilot studies
Brainstorming: Think through your proposal to be sure it is
logically sound, contains something original or out of the ordinary,
and includes work that will be a service to your research
community. Would you want someone else to get money to do the
project? (…answer should be ‘yes’).
Caucusing: Your research will ultimately be judged in the court
of scientific discourse, so it should not be conceived in a vacuum.
Talk to your colleagues about your plans. Would they want to do
the work themselves? Do they think your ideas are really novel?
Do not isolate yourself out of fear of competition or criticism.
Pilot studies: These are incredibly effective. They demonstrate
commitment and ability, illustrate expected products and
interpretations, and give you a test whether your original ideas are
really so hot. I suggest making them the center piece of every
proposal. They should not be used as an excuse to raise money for
studies you have already finished.
The care a feeding of your program officer
• Who are program officers? Accomplished academic
scientists spending ca. 5-10 years at NSF as a service to the
community and a change of pace from teaching and research.
Some like the work and stay longer. Treat them like colleagues.
• Who is my program officer? Find out as soon as
possible. You can check the web, but it is best to talk with close
colleagues about this persons’ background, likes, etc.
• How do I develop a relationship with him or her?
In lots of small ways. Seek him or her out at major meetings, to
talk about science and his/her interests. Call when you are
planning a proposal, at a quiet time of year. Do not rely on e-mail.
“casual Friday” in the D.C. offices
The care a feeding of your program officer
• What do they want to talk about? Program officers
are interested in you and your work, but not as much as you
are. Talk to them about things they will find interesting:
Their work at NSF; their scientific discipline; major efforts
at NSF, DOE or NASA that they might know about.
• What does my program officer want to do with
his/her program? This is key. Most seem to mentally
divide their programs into things they should support
because it is their job, and things they want to support
because they find them interesting. Learn which is which.
• What if I don’t want what they want? Continue to talk
to them, but don’t force proposals or mold your interests to
meet their goals.
Crunch-time
• Have completed brainstorming and pilot
studies ca. 1 month before submission.
• Don’t start committing words to the page any
later than 3-4 weeks before submission.
• Do the busy-work first. There is lots of it,
and it breaks writer’s block. This includes: CV,
list of current and pending support, list of
facilities and resources, most of your reference
list, and all or most of two sections of the project
description.
• Most programs have real deadlines; a few
(esp. instrument programs) are ‘soft’
• Administrators need ca. 2-3 days to process
your proposal
The Outline
• Results of prior support (ca. 2 pages). What did I do with the last check you wrote me?
Do not brag, ‘shingle’, pad or exaggerate. State the facts of your accomplishments.
• The problem (ca. 3 pages). Summarize the nature of the problem to be addressed, and
successes or failures of previous efforts. Why is this a good subject? How is it related to other
areas of science? What will change about your discipline of it is solved?
• The approach (ca. 3 pages). What is your angle? What are you going to do that
substantially solves or advances this problem?
• Pilot study (ca. 3 pages). Concretely demonstrate how your approach works by way of a
model study, ‘breadboard’ instrument, or comparable first pass through a problem.
• Work plan (ca. 3 pages). What, exactly, are you going to do? What is your time-table?
Do you need specimens or equipment; if so, where will they come from and when? What are
the anticipated products of your research? How will you know when you are done and
whether you succeeded? Identify and clearly state the risks or weaknesses of your proposal
(they are probably obvious to others anyway).
• Wider impact (1 page). What will you do that advances NSF’s goals of education,
diversity and advancement of the foundational infra-structure of your field?
The ‘cartoon-strip’ outline
• Roughly draft ca. 10-15 figures and 1-3 tables
that are relevant to your proposal, and write ca. 50100 word captions for each.
• Put them in order, as you would if you were
giving a talk on the subject.
• Pare them down until you have seven or eight
total figures and tables.
• Re-draft and re-size them so they look nice and
are big enough to read at arm’s length. Each
should cover ca. 1/4 - 1/3 of a page.
• Write the ‘hard-wired’ sections of your proposal
(results of prior support; statement of wider
impacts).
• You have now finished 1/3-1/2 of your proposal!
Write the main body of the text to follow the
logical flow of the figures and tables.
How much money do I ask for?
• Always ask for what you need.
• Make sure what you need is comparable to an average grant from
your division to a person of your seniority/group size.
• Common, reasonable requests:
1/3 to 2/3 of a major piece of equipment
100 % of graduate student salaries+benefits
50-100 % of postdoctoral salaries+benefits
ca. 15 % of your salary, but usually distributed over several grants
Travel for one meeting per person per year
Supplies and goo-gahs equal to ca. 10-25 % of the salaries
• Go for 3 years of support when you have a very strong case and
a proven track record.
• 1-year grants for things other than instruments are usually more
trouble than they are worth.
Seven deadly sins
• Proposal is incomplete (lacks some required section)
• Proposal lacks concrete, tangible goals
• Chosen approach does not speak directly to the problem
• No clear prediction of the differences between ‘positive’ and
‘negative’ results
• PI has no credibility with a problem or method
• Beating on a solved problem
• Using an out-dated or otherwise inferior approach
Five golden rules
• Never write a proposal unless you need the money
• Never ask for money to do something you don’t really value
• Set a limit on the number of proposals you will write per year,
and stick to it
• Do not get pulled into collaborative proposals unless you
believe in them
• Set goals that will provide a service to your field