Reader J asked  (In July! J, I am sorry for being so late with this response!):

If you wouldn’t mind a blog request:  now that I have survived the postdoc phase, got a job, have survived some teaching, written the boatload of grants … I got an NSF grant this winter (I’m still shocked, actually!) and although I know the basic idea is:  “do the work”, I could use some mentoring / blog posts on: the care and feeding of your grant.  How often do you contact your PO, and what do you discuss? Up until now, I’ve only discussed prospective Aims, and study section comments. What do you aim for for progress reports?

In my specific case, my institution’s grad program isn’t that great, and my grad student just dropped out to take a more lucrative job.  I’ve put out an ad for another postdoc immediately, since grad recruiting takes so long, but I feel like with NSF money I should be seeking to train grad students, and undergrads (have them) – is this a big problem, or should I be looking to be productive first and mission-oriented second?

First, congratulations on receiving an NSF grant!

I will answer based on my experiences, which are likely sort-of universal, but I assume there are some differences among directorates/divisions/programs.

First, hiring a postdoc when your budget specified students only: this should definitely be cleared with your program officer (PO), because, as of a few years ago, if you want a postdoc, you have to submit a postdoc mentoring plan, and it gets evaluated during peer review (you never did, because you originally didn’t plan for a postdoc). So it’s not just a simple matter of rebudgeting. Best-case (and likely) scenario, you contact your PO, they say no problem, just submit a postdoc mentoring plan, and that’s it. I have done the converse (budgeted for a postdoc, hired grad students when no good postdoc found) a couple of times with a couple of different agencies and it was never an issue. Indeed, the most important thing is to get research done, and if you have issues recruiting good students and can find a good postdoc, that’s a strong incentive for staffing change.

In general, NSF cares that you do good science with their funds and publish in good venues. They do not micromanage, so you don’t actually have to contact your program officer ever again in your life if you don’t want to, as long as you submit your annual reports on time. However, I recommend touching base with your program officer especially if you have some exciting data or a new high-profile paper, as that helps them look good internally, and also helps the agency make a case for their own budget in front of the congress. Some of the POs are rotators, so by the time your grant expires they will be gone; you can decide how that affects your attitude towards regular contact. But with permanent POs for sure, and I recommend not treating rotators any differently, touch base on occasion and simply ask — do they want to see your papers, how often? Basically, what can you do to make them and their program look good with their bosses? While with the NSF the panels are key, the PO is who picks the panel composition, who chooses who reviews your proposal, and who can pick you up from the pile of Recommended (but not Highly Recommended) and get you funded anyway. And it is possible (though rare) to get NSF proposals continuously renewed if you have been very productive based on this productivity alone; I have never had this happen, but I know some Greybeards who’ve had a continuous NSF grant for decades.

I have had a grant with another agency for years now (competitively renewed) and I always send my new papers to the PO, largely because I know he’s a huge science geek, really curious and passionate about the field and with a strong technical track record before becoming a PO, so I think he actually enjoys reading these. Whenever there is an editorial highlight, or a cover article, or anything notable with the work they funded, I most definitely let them know, because it is important and they use these highlights to lobby for more money and to keep getting funded.

What goes into an annual report? Well, those will be submitted through https://www.research.gov/ and I recommend going through the PI demo site.  I have had different POs request different level of detail for accomplishments, so I always prepare Major Activities and Accomplishments as separate files (with figures) and upload. There is a lot of information that gets entered in text boxes, most of it mundane. In principle, you can submit perfectly passable reports with minimum bling just by filling out the text boxes. I recommend asking the PO if they have use for a more detailed technical narrative of Major Activities or Accomplishments, or any other thoughts on the annual reports.

With the NSF, some POs are responsive and enthusiastic and welcome interaction with their PIs; others are hard to get a hold of, grumpy, and unresponsive. I have a suspicion that the first kind may slowly transform into the second kind as the stress and drudgery take their toll. But doesn’t it happen to the best of us? If you have a pleasant and responsive PO, enjoy interacting with them! But ask first what they need and expect — no harm in being explicit. They will likely appreciate it.

What say you, blogosphere?