Sunday, July 17, 2011


Eli is exchanging comments with Thoreau over at Unqualified Offerings. Poor fella got Goods on grant reviews. Today you need Very Goods and Excellents. Now the Rabett doesn't hit on every proposal, but he has had his share over the years and has read more than a few.

This is a first of a series of posts which will not tell you how to win a grant, but hopefully will show you what is needed not to lose one. It is US centric, but the broad themes are the same the world over unless your nephew is the Minister of Science, or Silvio Berlusconi's girlfriend. Eli will assume that your science is great but


That means read the grant manual. Yes, you need kick ass science, and yes, today paylines are under 10% and shrinking. Reviewers and program managers are looking reasons to differentiate between the 25% of the proposals they would give their inheritance to have thought of in order to separate out the 8% that will be funded.

Every call for proposal and grant manual has a list of hoops to jump through. They may not mean much to you, but they do to the program manager and agency. Just do it, but do it right. This is serious stuff.

So you read the call for proposal with highlighter in paw and make a list. And the call for proposal says that the proposal has to meet program goals which you go and download, and read and highlight, and there are then references to agency goals and you go read and highlight them.

Your proposal has to explain to the program and the reviewers how it will advance the goals of the funding agency. You don't need to write a lot, but you do have to show that you are aware of the issues. This is most important for EPA/DOD etc. and least for NSF, but even for NSF you have to discuss relevance to the program/division you are proposing to, and how (dread) your proposal will meet NSF goals for broader impacts (more about that later).

And of course, your proposal has to have all the requirements in the agency program guide.

RTFGM, highlight all the requirements and make a punch list.

Some of it is grade school. The number of lines per inch, number of characters per inch (no kerning allowed), margins, number of pages in each section and more, including allowed fonts. Program managers (well, some of them) need to triage proposals because they have so many of them to deal with.

There are suggestions, like use one column, yes, two columns looks neat and you might be able to squeeze in some extra text/figures but reading two columns on the screen is a royal pain and the reviewers will take it out on you. They will. Been there, seen that.

There are choices, paper,, or the agency web site for grant submission. Although it has gotten much better, if you have a choice, avoid Fastlane (NSF) and nspires (NASA) are better. Ask anyone who has not. Oh yes, don't do paper even if you are allowed to. The agency and the reviewers will take it out on you. It is a sign that you are not professional.

You have to register for all of these sites AND SO DOES YOUR UNIVERSITY. This takes time, esp. if your university is one of the NRHUs (Not Research Habituated Universities) and is not registered. If you think or even suspect that you will be submitting a proposal, REGISTER NOW and make sure your university is registered. This takes time.

The procedures are not always straightforward (for example, someone at Eli's place used a very awkward way of naming the organization at one of the agencies and it has never been changed, so if you don't know, you can't find it, or sometimes you only find it after screaming at the computer for a while. In general even if the Uni is registered, someone in the Sponsored Research Office is going to have to approve your registration. They may not notice that you have tried to register, so you should talk to them. That is a excellent time to discuss their procedures and what they can help with. It is also a great time to find out what internal forms and signatures you are going to need and how long it takes for each step with and without begging.

After you have registered (this is WAY before you are going to submit) put together a dummy version of a proposal. It doesn't have to make sense, the text can be your version of the quick brown fox, or some student lab report, or your thesis or whatever. The purpose is to give you a feel of the web site and how it uploads and checks the various parts of your proposal, how it handles such things as generating budgets, institutional information including DUNS, CFDA, congressional district (yes, the agencies want that so they can notify the congress critters that money is flowing into their districts) etc. If you don't know what some of these things are, your SRO does, or sadly, should, because many of them are asked for. You are going to need help navigating some of the forms, like the SF424 which is essentially the title page and a good reason to avoid You DON'T want to learn how to do this with the clock ticking down on your deadline.

Research universities (RHUs) have staff that do much of this for you, at NRHUs you are on your own, but even for folk at the RHUs being aware of such issues is useful.

Putting a proposal into the system does not submit it. There are two additional steps. First you have to submit it to your SRO (done on the web), and then they have to electronically sign it and submit to the agency. If you never submit the proposal it just sits there and no one but you knows about it. Oh yes, only registered organizations can submit, because the government requires that they qualify for reporting, audit, etc. There are organizations who exist as housing for proposers at the cost of a percentage of the grant.

The next step is beg to see a winning proposal or better proposals. Ask folk you know, take a look at the agency web site, ask the program manager whom to approach, etc. Winning proposals are open to the public, but as a practical matter, the agency will a) redact a whole lot of stuff you want to know and b) not be happy with you for asking because they have to spend time redacting, so it is much better to get one from a friend or colleague. Remember, you are applying to a program, you want a winning example FROM THAT PROGRAM.

Tomorrow, convince the reviewers that you are a pro.


Pinko Punko said... has improved, while fastlane has not been updated in awhile. Also "very goods" won't cut it anymore. We're getting into insanity territory now.

amoeba said...

"...unless your nephew is the Minister of Science, or Silvio Berlusconi's girlfriend."

I know that Eli is much smarter than me, but this left me really confused.

Thomas Palm said...

Makes you wonder how much research could be done if scientists didn't have to spend half their time writing and reviewing grant proposals.

Deech56 said...

And there are opportunities to be awarded contracts as wel (if I were younger or retired, I would go into the consulting business). First rule: if there are mandatory criteria, the description for said criteria should be no deeper than 5 pages into the proposal. Answer each aspect of the ask, but don't get to bogged down in the detail. You an always provide more information during negotiation.

Mark said...

PP, the last proposal I submitted received all "very goods" and was declined. The program officer told me he would have liked to have funded it, but nowadays you pretty much need all "excellents".

EliRabett said...

"...unless your nephew is the Minister of Science, or Silvio Berlusconi's girlfriend."

they provide set asides

Pinko Punko said...

Mark is correct. The scale to me seems 0-10% Excellent, 11-25% Very Good, 25-40% Good, 41-70% Fair, 71-100% Poor. Something like that, although with score compression, I would be surprised if "Very Goods" were a whole quartile or larger. There is just no money.

John said...

Right now, the whole Fed govt is on Continuing Resolution, so funding agencies have a very hard time finding the funding for new starts (i.e., funding a researcher who is not already funded).

Thanks to Eli for mentioning all the possible things that can go wrong when writing a proposal. When I submitted my first proposal to NSF way back in 1981, NSF accepted paper copies and wanted 15 photocopies. So I made 15 photocopies, and a few more for various university offices. Then I went to the chief bean counter, who found a minor error in the budget. I had to back and re-do the budget, photocopy the budget pages, tear open every photocopy, and substitute the revised budget, and re-staple. Then off to Federal Express to submit the budget.
I learned a lesson from the process: get the budget numbers blessed by the bean counter in the University Research Office BEFORE making any photocopies.

After the Internet became popular, proposals were submitted to NSF over the Internet (via Fastlane if you're lucky). Then NSF personnel made a bunch of photocopies and distributed the copies to its various offices. So the upshot was that the number of photocopies remained the same, but Fed Ex laid off a few drivers.