Machinations

I haven’t been posting in a while because I’ve been busy with preparing NSF grant proposals (which have shifted their deadlines this year from spring to December).  In the interest of getting out a post (and procrastinating), I’m going to write today about my great passion as a researcher: grant writing.  Seriously, I do sometimes enjoy the opportunity to plan out a research agenda for the next 3-5 years, and always the proposals I write do inform the research I will do in the future.  Also, so far I’ve also been really lucky with  grants, with a hit rate of about 60% (we’ll see how it goes this year now that we have a toddler around the house).  When your grants are funded, the grant writing process is much more fun!

Here are some of the tricks that I’ve learned for good grants.  Mostly these are from my experiences with NSF grants but they have been useful for other agencies.  I’d love to hear about some other ideas in the comments section.

The Introduction:

• Nail Motivation: Start with the problem you are trying to solve.  Show that solving this problem has big implications, hopefully outside your own narrow area of work; and that your research will take significant steps towards solving the problem
• Nail Novelty: What is it about your approach that will cause you to succeed where others have failed?  What unique ideas/tools/techniques do you bring to the problem?
• Expose a gap: As early as possible in the proposal, describe a gap in current research
• Don’t repeat: Don’t repeat text from the grant summary in the intro of your grant!  This risks starting a bad precedent of the reader skipping over parts of your grant.
• Pull in the reader. Write a first sentence for both the summary and intro that is controversial, asks an interesting question, evokes strong emotions (hopefully not disgust :), etc.

The body:

• Less is More: Don’t tack on ideas or problems that you don’t really understand.  The reviewers will uncover your ignorance and your grant will be skewered.  The main constraint on conveying ideas is not the page length, but rather  reviewers time.  Expect a reviewer to spend about 30 minutes on your grant.  You can really only get across a couple of clear ideas in this time.  I know this from experience – whole sections of my grants have been misunderstood by reviewers in the past because I tried to cram too much in.  Make it clear what the grant will not cover as well as what it will cover.
• Pictures are crucial: Spend as much time putting in 2-3 useful figures as you would writing a section of the grant
• Related Work is crucial: You’re proposal should be like an exciting conversation with the research community.  Thread related work throughout your proposal – be generous with praise for great ideas, and make it clear how your efforts fit into the broader scheme.  If you make the related work look exciting, your own research will look more exciting.  Don’t miss anything that is remotely relevant to your work.  A common way to have a grant shot down is a review like: “This proposal ignores results  from the work of researcher X”.
• Technical depth: a portion of your proposal (maybe a third?) can and should be addressed to the experts.  This is the place to include proofs, equations and enough technical details to convince the experts you have a good idea.
• Try to give something to the reader:  teach the reader something new, include an interesting story or analogy, describe a cool problem.  The reader will appreciate anything that breaks the monotony of reading grant proposals.

Some ideas about the process:

• Give a talk: I have found it useful to give a talk on preliminary results and future ideas as part of the proposal writing process.  It works well to do this for both an expert audience and a non-expert audience.  See what questions and ideas you get and use these to improve the proposal.
• Become your own worse enemy:  you absolutely need to anticipate the 2 or 3 most devastating criticisms of your grant and then address those in the proposal
• Get feedback: Get an expert and a non-expert to spend half an hour with the grant and give quick feedback,
• Be intellectually mature: Don’t pretend you’ll solve every problem.  Give credit where it is due.  Be realistic about  the research time line and the limitations of your technique.
• Try to write an hour a day on the grant to keep up momentum.
• Develop a style: I like to use a sort of dialect style where I describe problems and then solutions, or questions and then answers.  This usually determines the broad outline of my proposal.  However, there are many different writing styles and you should figure out what works best for you.