Proposals and the National Science Foundation

For the last week or so, I have been attempting to finish a proposal that I would like to submit to the National Science Foundation (NSF). I have been having a lot of problems with this because the University of Michigan has been getting a large number of proposals returned from NSF without review due to not following NSF’s Grant Proposal Guide precisely.  Here are some examples:

On a person’s CV, you have to list all of your collaborators in the last 4 years. In the past, to make life easier for me, I have listed these collaborators in chronological order, so that I can bump the last people off the last and add more at the top.  It seems like a lot of people do this, because NSF has started to return proposals when this list is not in alphabetical order.  Oh, and it clearly states that you have to indicate the number of collaborators, and not just list them. Count correctly!

One of the first reasons that we started to get returned proposals was the use of “et al.” in references.  If you have paper with more than, say, 15 authors, it is common practice to list the first author and “et al.” for the rest of the authors.  NSF returns proposals without review for this now. List those authors people! Even if there are 100s of them!

It used to be that you would discuss what your collaborators (unfunded people on your proposal) would be offering your research inside the proposal.  NSF now wants this to be discussed in an additional section called “Facilities”, where you normal put a cut and paste statement about how UM provides desks and computers and stuff like that.  Now, people are facilities too.

More CV fun: I had a proposal returned without review because I accidentally listed 11 papers on my CV in one section instead of having 2 distinct sections (most relevant and more recent) with a maximum of 5 paper in each section.  And, the publications are now called “products”, and you can list web pages and other “products”.  Luckily, the program manager allowed me to resubmit my proposal after making this 5 minute alteration. You now have to state explicitly how many Ph.D. advisors you had, even though you list them (and 90% of the people have only one, so you put something like: Ph.D. Advisor (1): Fred Smith, EMU). Same for post doc advisors (also listed, most likely, also one). Same for students that you have mentored (also listed). And post docs that you have mentored (also listed). Can people not count? To one? To two?  What is the point of putting the number of students, when they are listed?

I just got an e-mail yesterday that stated that someone’s proposal was returned because their section on “Broader Impacts” was labeled “Broader Impacts” and not the required “Broader Impacts of the Proposed Work”.  I shit you not.

So, what is going on?

It used to be that you got a proposal to review and you reviewed it based on what was included in the proposal package.  If they didn’t include something important (like the number of students that they mentored?), you dinged them on it and gave them a lower score.  The idea is that the proposal is reviewed on the merits.  Funny that.

Now, it seems like the NSF is attempting to make every proposal uniform. It must contain the same bits of information as every other proposal.  In exactly the same format. Exactly. I can see two reasons for this:

1. You can then say that there is a “level playing field”.  Since everyone has to provide exactly the same information in exactly the same format, then all proposals can be judged against each other in the same way.  We can then use an unbiased method for proposal evaluation and people who were doing really well will be more on par with people who were not doing nearly as well.

2. Since proposal pressure is becoming significantly larger due to budget constraints and rising costs of graduate students and such, NSF needs excuses to return proposals without straining the system.  So, if they return 10% of proposals without review, it helps the system a bit.

Both of these rationale are completely flawed.

There is no leveling of the playing field, since the meat of the proposal can be radically different.  CVs matter a tiny bit, but only if you don’t know the person.  Let them list 20 papers.  Let them list 5 papers.  It doesn’t matter.  What matters is the idea of the proposal.  Is this a good idea?  Does it investigate an interesting idea?  Is the methodology interesting? practical? boring? These are things that you can’t regulate.  If NSF wanted to really level the playing field, they would offer workshops on how to write better proposals.

(As an aside, NSF program managers are completely awesome.  When you are on a review panel and you know that the person just doesn’t know how to write a proposal, you can write a review that describes the proposal writing process and give advise and on all sorts of stuff.  At NASA, you are, in no way, allowed to do this. Not that the NASA program managers don’t want you to, they just can’t.  Once again, let me repeat, NSF program managers are awesome.  This shit is coming from higher up.)

By returning 10% (more?) of the proposals without review, all NSF is doing is forcing people to adapt to the new system and wasting time.  Eventually, everyone will get to that new system, crossing all of the “t”s and dotting all of the “i”s.  The only way that they can keep up the 10% (or more) return rate is if they keep changing the requirements.  This just then forces more time spent on making sure you know the ever changing rules, and is a complete waste of time.  Making science even less efficient than it is now.  Which is insane.

What would be a MUCH better way to reduce proposal pressure would be to limit proposals to two pages. The 2-page proposal gets (quickly) reviewed by two outside reviewers.  No budgets, no “Post doc mentoring” section, no “Data management” section, no institutional endorsement, etc etc. If they both say “yes”, then you are allowed to submit a full 15 page proposal.  This would dramatically reduce the number of real proposals.  It would help the researchers, since they would have to pass through this gate in order to actually submit a real proposal.  Rates of winning a real proposal would go up dramatically, since there would be significantly less (real) proposals entering the system. Reviewing these two pagers would be trivial, since it would only take a few minutes to read through it and say “yeah, this is interesting” or “no, this sucks”. One paragraph of review. Boom. Done.

Maybe I will pitch this idea to NSF people. Think it will work?

Be the change!

Advertisements

About aaronridley

Professor at the University of Michigan, Department of Climate and Space Science and Engineering.
This entry was posted in Uncategorized. Bookmark the permalink.

One Response to Proposals and the National Science Foundation

  1. aaronridley says:

    Just received this morning from NSF: “Beginning March 18, 2013, FastLane will enforce submission of required sections of full proposals prior to proposal submission. If you are responding to program solicitations that may deviate from GPG instructions, conferences, symposia or workshops or to international travel grants where the submission instructions do not require one of the sections to be provided, you will still need to enter text or upload a document in that section of the proposal that states, “Not Applicable.” For more information, please refer to the Advisories on fastlane.nsf.gov or refer to http://www.nsf.gov/bfa/dias/policy/autocompliance.jsp

    autocompliance.jsp says it all – review by Java!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s