NSF GRFP 2012-2013

[InquilineKea]

Is it a good idea to make my personal statement 80% focused towards Broader Impacts? (that actually could potentially be one of the strongest points in my favor given my significant Internet-based outreach).

I've read a number of NSF Fellowship essays now, and a lot of them talk about the applicant's personal development through college. I'm not sure if those points are especially informative when it comes to addressing "Intellectual Merit" or "Broader Impacts", and most of them don't sound especially memorable, but maybe there's a reason why so many of them talk about personal development rather than "Broader Impacts"? A lot of winners have personal statements that also have a lot of overlap with their previous research too..

Also - I have ADD and the form requires me to disclose disability. I'm not sure whether or not I should do it... It could really help with broader impacts (and explain my GPA during my first couple of years - I eventually got medication mid-way through college and my grades shot up), but it does carry A LOT of stigma with it.

Edited September 29, 2012 by InquilineKea

[vertices]

A lot of people address personal development because that's what the prompt asks and they are trying to address all of the questions in the prompt. It's best if you can address them in a way that bolsters your Intellectual Merit and/or Broader Impacts. While these are the two review criteria, they're not so rigid, so sometimes personal experience can demonstrate them in ways that aren't directly tied to the short summaries of what they are. For example, showing your leadership potential tends to be very important. Though that's not directly emphasized in the Intellectual Merit/Broader Impact explanations, it makes sense because leadership is an extremely useful characteristic in being able to be transformative or have broad impacts. Also, I think often the personal development can support how/why an applicant's Intellectual Merit/Broader Impacts are believable.

I suggest you address all the questions, but if you can do so in a readable way while also being able to devote 80% of your personal statement to Broader Impacts, that would be great. Probably you can weave your personal narrative through your Broader Impacts activities to show your passion and skill in your area of research and how it already has positive influence on the community at large.

User irugga seems to have had some positive experience addressing a disability. I see you've probably read the posts already, but I'll link the posts for anyone else looking for similar advice:

http://forum.thegradcafe.com/user/68719-irugga/page__tab__posts

Several people have reported getting comments about their GPA in their reviews, so it would be good to have some explanation of any GPA hiccups.

[InquilineKea]

Wow I see - thanks so much for all your awesome answers! irugga is *really* inspiring - I just sent a PM.

Another question: If you have an interdisciplinary proposal that gets reviewed by more than one panel, then will you get more reviewer comments than people who have stuck strictly in one subdiscipline?

Edited September 29, 2012 by InquilineKea

[warbrain]

Just to double check, we can only use 12 point font for our essays? I've looked at some other essays online and they seem to use much smaller fonts. And I'm pretty quickly running out of room on my essays, and would definitely prefer to use a smaller font size.

[vertices]

warbrain - Definitely stick with the stated rules. Some of the essays online might be from before that rule was in place. I've seen posters claim to get away with 11.8pt, but I don't think it's worth the risk. One poster got a notice of disqualification for an artifact/glitch in the margin picked up by the software they use to validate format. I'm not sure that got that sorted out.

It's painful to cut, but also a useful exercise. I ended up cutting a lot of things down to essentials and entirely omitting some points. For me at least, I think it helped make my writing clearer and my message stronger. I suggest you finish your first drafts and then go through sentence by sentence and ask yourself what the sentence is trying to say, whether you can say it more succinctly, whether that point is addressed elsewhere in your essays, and how important that point is compared to everything else. Good luck and good writing!

InquilineKea - Most interdisciplinary proposals will only be reviewed by one panel, just like the single sub-discipline proposals. The application asks for a listing of all fields represented in the proposal. The one listed first determines which panel will review it and what deadline applies. While the panel can transfer the application to another ( http://www.nsfgrfp.org/how_to_apply/choosing_a_primary_field ), I would guess that having reviewers from multiple panels doesn't occur too often because the reviewers in any panel are experienced in reviewing interdisciplinary work.

Each application gets two reviewers minimum. The top ~35% of applications get a third reviewer as does any application where the two initial reviews diverged notably. With those two common reasons, it would be hard to attribute a third review to multiple panels instead.

[guttata]

Just to double check, we can only use 12 point font for our essays? I've looked at some other essays online and they seem to use much smaller fonts. And I'm pretty quickly running out of room on my essays, and would definitely prefer to use a smaller font size.

Just to add, the one exception is for citations and conference proceedings/publications, which you CAN put in a 10pt font. That does give you a few more lines for the main essay.

You can also play around with a few tricks. I became very adept at revising paragraphs that ended with 1-4 words on a new line, so that almost all of mine end more or less flush with the right side of the page. For one or two essays I also used 5 taps of the spacebar instead of the tab button. It adds up.

The fine details of citations aren't always incredibly important. Instead of doing a full citation with spaces and indents such as:

Smith, DJ, TF Jones, LS Jones. 2012. Birds fly and have feathers. Journal of Birds, 34: 123-129,

Smith DJ, TF Jones. 2011. Birds are animals. Journal of Birds, 33: 99-101

I use an extremely truncated version in a giant block like so:

1. Smith et al. 2012. J. Birds 34:123-9 2. Smith and Jones. 2011. J. Birds 33: 99-101

This citation style reduced 2 citations to less than the length of 1 full citation. Especially on papers with long titles, this adds up. Intext citations are a superscript number. Didn't get dinged for it at all last year and I've seen others do similar things.

Edited October 7, 2012 by guttata

[guttata]

Another question: If you have an interdisciplinary proposal that gets reviewed by more than one panel, then will you get more reviewer comments than people who have stuck strictly in one subdiscipline?

From the link that vertices provided (Emphasis mine):

Choosing an appropriate field of study is a very important consideration for your GRFP application because it determines which panel will evaluate your application.

Each primary field is associated with a specific panel, and all applications with that primary field designation are assigned to the panel for that field.

So, from this statement - where only panel is mentioned, never panels - ONLY your primary designation matters, unless the panel members think you screwed up. Honestly, I would think that if the reviewers have to send your app to another panel you're going to get dinged for that. Why would they fund someone who can't even tell where their research falls? On the flip side, I also think you would really have to get your primary field wrong to have them send it off.

Edited October 7, 2012 by guttata

[Pitangus]

I use an extremely truncated version in a giant block like so:

This citation style reduced 2 citations to less than the length of 1 full citation. Especially on papers with long titles, this adds up. Intext citations are a superscript number. Didn't get dinged for it at all last year and I've seen others do similar things.

I did the same thing: I modeled my citation format after a "Brief Communications" paper in Nature, and put the references in a block. Also, I used 10 pt font for the references. I was not disqualified and I did not receive comments about my reference list, so I'm assuming it was not a problem.

Edited October 7, 2012 by Pitangus

[vertices]

I did the same thing: I modeled my citation format after a "Brief Communications" paper in Nature, and put the references in a block. Also, I used 10 pt font for the references. I was not disqualified and I did not receive comments about my reference list, so I'm assuming it was not a problem.

I also did 10pt brief/block references for the proposal and have also seen several other winners do it too. For the Research Experience, I put a line at the end stating the references referred to the "Honors and Publications" section of the application, following what previous winner Kevin Karsch had done: https://wiki.engr.illinois.edu/download/attachments/160301097/karsch1_NSFGRF09-Previous.pdf?version=1&modificationDate=1283980287000

[Alext182]

Has anyone found essay examples for fields within the social sciences? I ask because it seems that the Broader Impacts of social science research is typically more abstract than research in the STEM fields. If any previous award winners from the social sciences read this, please feel free to PM me if you prefer. And thank you to the posters about the most efficient way to do references.

[arrowtotheknee]

@Alext182 - Why would the broader impacts be more abstract? Could you give an example of what you mean? The example proposals I have seen often have more explicit and applied impacts than those in physical sciences. If you are having difficulty defining the Broader Impacts, you may need to reevaluate your research question.

For sample essays, I would ask your department and/or advisor. It's also possible that your school has workshops about how to win the award -- sometimes these go over example essays, too.

[Alext182]

@arrowtotheknee - Broadly speaking, fields that focus on the relationship between social institutions and individuals typically differ from the STEM fields, because they are more abstract and less focused on finding THE solution (eg sociology, political science, communications, etc). For example, scholars in these fields are less likely to conclude their studies with concrete solutions to the problems they may study, such as inequality.

I am not having difficulty defining my broader impacts, rather, I am having trouble identifying what the standards are for social sciences. The numerous examples I have found are typically related to creating/identifying solutions related to scientific problems (eg diseases, environmental problems, city infrastructure, computer programming flaws). These STEM studies make up the bulk of the awards, as they should, but I find it odd that it is difficult to find examples of social science winners online.

Thanks for the advice. I've found examples online, contacting past winners, and my university.

[arrowtotheknee]

From what I've seen, the standards are similar across all fields: will the proposed project help us better understand some interesting question? Can the candidate define an interesting question? Will answering the question have any small (ie to subjects, the subjects' community) and large (science community, society, etc) impacts? The social science examples I've read have been more or less concrete and targeted questions, typically with quantitative methodologies.

[SensLu]

I am really curious about if anyone was able to overcome getting this award with a low GPA at around 3.0 or below and if they explained a low GPA in their personal statement. My guess is no one with a low GPA would bother applying. However, I plan to apply anyways for the experience and the fact that I don't feel as bad since I have a lot of experience leading research projects, presentations at top conferences, doing outreach, excellent letters of recommendation and according to comments made about this award in the past, demographics seem to be on my side.

[Usmivka]

I am really curious about if anyone was able to overcome getting this award with a low GPA at around 3.0 or below...My guess is no one with a low GPA would bother applying.

I suspect you are correct.

My impression is that what you write and propose is what is being evaluated, however, not your academic record. Good science doesn't have to come from good students, different skill sets are involved in being a researcher and a student.

Edited October 16, 2012 by Usmivka

[Jema]

My impression is that what you write and propose is what is being evaluated, however, not your academic record.

I beg to differ with Usmivka's assertion. Unlike NDSEG and the DoE fellowships, GRFP does not give much weight to the merit of your proposal. Its what you have done till date that counts like research, academic achievements etc and yes, you need to have an acceptable GPA. It need not be perfect though ( like a 4.0 GPA ). There is also the caveat that a low GPA from a good school with no grade inflation (like Caltech) may be acceptable. They need to be convinced that you have the aptitude to do research and is in for the long haul. Also unlike the grad school admission process, I haven't heard about many low GPA's winning NSF fellowships. An application with a good proposal with less than steller credentials has less chance to be selected than the other way around. But if coba11 has some good research to back the application, along with good LOR's, its worth giving a shot. I've felt Lor's are key in situations where something in your app pulls you down. You can also mention in the personal statement about your low GPA especially if you've had a steep recovery after a few bad semesters.

Edited October 16, 2012 by Jema

[guttata]

There are multiple comments in past years from students with relatively low GPAs having success winning the award. The common thread in most cases seems to be that at SOME point there were strong semesters - these weren't students getting 2.8s across the board. There was a semester with an illness, or a family problem, or something, that caused the dip or that the student figured and recovered from. The key is to note that and what you've done to remedy the problem and show how you grew from it. Off the top of my head, one student I know had ~3.2 (an acceptable GPA but maybe a bit low for most applying for the GRF) but won the award. He spent a good deal of his application addressing his dyslexia and time spent bouncing between majors (declared 3 different majors in 3 different fields during 4 years, IIRC) as contributing factors.

In short, it's not impossible, but you do need to spin things the right way.

[Usmivka]

I beg to differ with Usmivka's assertion. Unlike NDSEG and the DoE fellowships, GRFP does not give much weight to the merit of your proposal.

I made that statement because the reviews for my accepted proposal focused almost exclusively on the merit my proposed work with no comments about my academic background or prior work. That isn't to say they didn't look at those aspects of my application, but they certainly didn't comment on them!

Edited October 16, 2012 by Usmivka

[Jema]

The idea behind the research proposal is to convince NSF that you can think through to a solution for a problem. My understanding is that there are points for each part of the application. That's why sometimes people do get good to excellent reviews from all three reviewers but do not get a fellowship. After all they can only award so many from a particular field even though 2000 total fellowships a year may sound a lot. So the idea is to project oneself as better than the other applicants in the same pool.

[sheet_music]

Hello all,

I have a question regarding recommendation letters. I'm a second-year student and planning to have one rec letter from my advisor, and one rec letter from my unofficial co-advisor. For my third letter, I have two options: (1) my undergrad advisor, with whom I published a conference paper, although I haven't really spoken with her since last year's NSF GRFP deadline, and (2) a faculty in the CS department, who is the director of an outreach program for which I am the graduate student lead.

One of my reviewers last year complained about how I wasn't doing any outreach at the graduate level (although I was volunteering with two separate organizations and wrote two separate paragraphs about it...). I'm hoping that by asking "letter writer option (2)", I can make it absolutely clear that I both do outreach, and have a leadership role in these outreach organizations. However, the description of rec letters on the NSF website seems very research focused.

I'm wondering whether it is worth having a letter from my undergrad advisor as a second year grad student, or if it would be better to have a letter completely devoted to outreach/broader impacts.

What are your thoughts?

[Jema]

@sheet_music, You absolutely need to get an LOR from your undergrad advisor. There is no two ways about that. You can always write about your outreach activities in your personal statement mentioning the name of the CS faculty you are working with and the kind of work you are doing.

[Usmivka]

I'm wondering whether it is worth having a letter from my undergrad advisor as a second year grad student, or if it would be better to have a letter completely devoted to outreach/broader impacts.

I would use the undergrad adviser since you are published with them, but drop either your actual adviser or coadviser in favor of the outreach reference (I don't know how much weight a letter from your graduate adviser is given since he/she may have a significant financial incentive in getting you funded, program structure dependent, so I'd go with whoever can write you a better rec., but then again a reviewer could view that as odd).

Regardless, here is why I think you need that outreach focused letter, or emphasis on that point from your other references: One of the only two specified judgement criteria is broader impacts. "Contributing to general scientific knowledge" at a conference or in a paper doesn't really cut it on its own anymore. Your outreach is a big deal given NSFs emphasis on community outreach, and deserves just as much recognition as your research accomplishments.

I had a letter from a faculty member who only knew me in an outreach context, and I think it contributed to my acceptance. Certainly reviewers commented favorably on the "continued outreach" component of my proposed research. I felt safe giving one letter to my outreach person because I knew the other two letters would be good as well--I asked each writer to focus on specific things that NSF was looking for, based on their written evaluation guidelines. If you feel like your other letters might be weaker, then this could be more of a risk.

Edited October 17, 2012 by Usmivka

[Jema]

You need to select the best people who could write you a strong letter. But if you drop your graduate or your undergraduate advisor from your list, that would raise a red flag. The best way would be to get your undergrad and grad advisor to write about your outreach activities and other things that contribute to the broader impacts criteria. Try to get your graduate advisor to talk about the broader impacts of the research work you are doing currently (which I assume is the same as your proposal). If you cannot get them to do so , then it is imperative to get the third letter from the CS faculty you've talked about, but have you just started working with him? Can he write in detail about what you are doing/going to do?

[dendy]

Is it kosher to list papers that haven't been published yet? I have one accepted for review and one under revision. Would it also be ok to cite these in my essays? Can I just include all the standard citation info + (under review) or ?? And would you bother to list an unpublished honors thesis?

[agarcia59]

Is it kosher to list papers that haven't been published yet?

I've seen a few winning papers listing their submitted (under review) publications. I would suggest doing so.