Saturday, May 22, 2010

NSF CAREER Tips

It is mid-May, so there are 2 months left till mid-July. But, I am not counting down to vacation: mid-July is when NSF CAREER grant proposals are due. Therefore, around mid-May is when my junior colleagues usually start to write (or think about starting to write) their proposals, and I start getting a lot of questions about my experience (I got a CAREER award several years ago). In this post, I will try to provide, as much as I can, a discipline-independent set of suggestions for writing this type of grant, based on my experiences.

NSF CAREER awards are 5-year grants that are meant to enable promising young faculty to develop their research programs. Depending on the NSF directorate you plan on applying to, these awards are capped at either $400 K or $500 K.
Depending on your school's overhead and your lab needs, this usually won't cover much more than lab supplies, 1 student, and 1 summer month per year. But, the prestige associated with CAREER is significant, and getting it is a valuable (although certainly not necessary) ingredient in securing tenure.

1. Getting Informed

As with all grants, contacting the program manager (PM) is extremely important. Typically, not all program managers in your division will be involved in the CAREER program, so your regular PM may not have anything to do with it. So first find out who the contact PM is for the CAREER program, and contact him/her early on, before you start writing. Now is early enough that they may not be entirely swamped yet.

Also, some divisions utilize mail-in review + panels, some only panels (I don't think any do mail-in only). In some divisions, CAREER panels are composed of senior prominent faculty, in others you can have recent junior awardees. The composition of the panel and the PM may be significantly different than your regular program's PM and panels, so get informed early.
In my experience, you may need to be persistent in getting a hold of the PM, but this is very important.

2. Scope

In my opinion (I'm not alone in this), CAREER grants are different in scope than regular NSF grants. They are painted with broader strokes, you are allowed to dream bigger, more far-reaching dreams in the CAREER than a regular 3-year grant.
After all, CAREER is supposed to establish a foundation for a lifetime of continued achievement (paraphrased from how I remember the solicitation). I interpret this as the following: what is it that you would most like to do in the next 10 years? So write a proposal on the first 5 years of THAT project. A project of too narrow a scope is not appropriate for CAREER program. So if you have a nifty idea that potentially opens a new niche subarea where you will be a star, that's the type of project you want for CAREER. Once you have the idea formulated, run it by the PM. If you have a few and don't know which oneto write about, run it by the PM. Helpful ones will say "This one is too narrow, this one is appropriate."

3. Readability

As always with panel reviews, the writing style should be such that a panelist reading your proposal in the middle of the night before the review, after having travelled to Arlington all day and read several more proposals on the way, gets a clear idea of what it is that you want to do, and why it's exciting and important. In CAREER proposals, you are more costrained in space than with standard grants proposals: there is a 15-page limit as usual, but you ought to devote a much larger chunk to Broader Impact (about 5 pages, more about this below), so count on having 10 pages to demonstrate the vision for your career development, and describe the idea and exectution for this particular proposal and how it helps your long-term career vision. This means you may not be able to get excessively technical. Readability is key. I am personally also partial to short titles, but people vary on this.

4. Broader Impact (BI)

This part is extremely important for CAREER grants. As I said above, while the BI for a standard NSF grant would be about 2 pages, for CAREER it's more typically around 5. Many people think "education & outreach" when BI is mentioned, and while they are not synonymous, that is what I will start off with.

For a standard grant, activities such as inclusion of undergrads in research and plugging into your institution's existing initiatives on broadening participation may be sufficient. (Saying you'll advise grad students and disseminate results in journals doesn't cut it even for a standard grant; it's implicit you will do that anyways).

For CAREER, you have to go beyond and show some creativity. You don't want the education part to look as an afterthought, but as a natural outcome of your research. For instance, introducing a new course or augmenting an existing one with your proposed work would be a nice ingredient of your BI. Also, you may want to interface with local teachers to create K-12 modules that relate to your resaearch. Try to be creative in whom you target: K-12, community in the broader sense? Regional, statewide, or nationwide? Specific adult age groups? Specific community organizations (neighbourhood associations, churches)? Will you use web resources to access your audience, or perhaps newspapers, radio, or TV instead? Have you made connections with news outlets, schools, museums? Be specific in your plans.

As I said, your broader impact need not be limited to education and outreach. For instance, if you propose to cure the common cold or male-pattern baldness, then you should definitely emphasize the societal impact that you work will have and how you can help maximize it through utilizing existing institutional, regional, national, and NSF-funded resources.

Broadening participation is another important aspect of BI: is there a dearth of women or ethnic minorities in your field? Can you find creative ways of targeting any of these groups through research-related initiatives? Are there ethnic groups indigenous to your region, where you could make an impact? If there are already diversity programs (NSF-funded or not) on your campus, explore how you can leverage them in your BI.

In general, showing you have thought how you'll support your BI activities is important, so if needed budget for them in your CAREER grant. However, be aware that certain components of BI are eligible for extra funding: every year you can apply for research experience for undergrads (REU) and research experience for teachers (RET) supplemental funds to support participation of an undergrad or a teacher in the lab over the summer. There are also REU site grants, but these are something different. Your institution may or may not have an REU site, but you are eligible for REU/RET supplements to your own NSF grants, and CAREER is no exception. How hard/easy it is to get an REU or RET supplement depends on NSF division, in some it's a virtual guarantee; your PM can tell you with how much certainty you can count on these supplemental funds. If they are easy to get in your division, I recommend you mention in your grant that you will be seeking these supplements in the coming years, rather than have to budget for them in the CAREER grant directly.

Propose what you would actually enjoy to do for BI. People vary widely in how much time they are willing to invest in BI activities, so don't propose something you would hate doing, because it shows when you write.

Lastly, BI should enhance and complement your awesome research, not be a crutch. A great BI will not compensate for a less-than-stellar research project.

5. Letters of Support

CAREER is a single-investigator grant, so you cannot have co-PIs. However, it is quite likely that you need other people's expertise or the use of someone else's equipment for some part of the project. In that case, every such person should provide a letter of support (where they say they are excited about working with you, but need no money and are separately funded); these letters are included as part of your application under "supplementary documents".

I recommend including letters of support for your BI activities as well (e.g. from the principal of the school if you will do K-12 outreach, from organizations/programs with which you will interface for community outreach or to recruit minority students). Letters of suport show that you have thought about these initiatives seriously and made connections in advance.

Good luck!

(Check out some of the excellent additional tips among comments.)

5 comments:

prodigal academic said...

Nice post--I haven't seen this information anywhere else, so thank you for posting it!

Anonymous said...

Other tips from a recent CAREER awardee:
1. Find out what type of Broader Impact your specific review panel will be looking for. Some panels prefer proposals saying you'll develop a new class and have some undergrads/high school students/teachers in the lab; others will only like projects that are larger and show creativity. Many sr. faculty have told me that they prefer the first option because they feel it is more realistic. However, my specific panel did not like this approach the first year I applied. So the second year I applied, I updated the BI with a more creative plan and was funded.
2. Be certain to explicitly detail how you will evaluate the impact and effectiveness of your BI.
3. Ask as many people as possible to read your summary page. This is the most important part of the whole grant. It needs to capture the reader's attention and make it clear why your proposed work is important and interesting and why you are the right person to do it.

GMP said...

prodigal academic and Anon 7:26, thanks for the comments!
Anon, excellent points -- I agree completely!

Odyssey said...

I recommend including letters of support for your BI activities as well (e.g. from the principal of the school if you will do K-12 outreach, from organizations/programs with which you will interface for community outreach or to recruit minority students). Letters of suport show that you have thought about these initiatives seriously and made connections in advance.

As someone who has reviewed CAREER proposals I can tell you that these are essential. SImply proposing some creative BI's may not be sufficient. You need evidence that you have started working towards implementing them.

As far as standard NSF research grants go, the old "having undergrads in the lab for your BI's" approach is rapidly becoming insufficient. Program officers are pushing the panels to be more critical of the BI's.

GMP said...

Thanks, Odyssey, great input!