Graduate Fellows Discuss Fellowship Experiences

The following three NSF Graduate Research Fellows and an NIH F31 Fellow responded to questions about their fellowships for Amstat News. This is a companion piece to this October Amstat News article, “Fellowships for Graduate Study.”

Frederick Campbell is a graduate student at Rice University and is interested in optimization and statistical machine learning with applications in medicine. He was awarded the NSF graduate fellowship in 2012.

1. Please describe your approach to preparing and writing your application.
   The NSF GRFP application requires three letters of recommendation and three essays that communicate the intellectual merit and broader impact of your work. For the recommendations, I reached out to the professors I had strong relationships with as early as possible. Everyone I reached out to was incredibly supportive, and I am tremendously appreciative of their time! For the essays, I described my undergraduate work and the work I began at Rice. My previous work has applications in education and medicine, which made it easy to describe its broader impact.
2. Did you work with a faculty member on your application? If so, how?
   Throughout the application process, Genevera Allen was very helpful. Beyond advising several projects that I was working on Professor Allen encouraged me to apply for the fellowship and ensured I wrote many, many drafts of the application. All my writing is the product of lots and lots of rewriting, and Professor Allen was available to provide feedback at nearly every iteration. I don’t think I could have won the award if Professor Allen wasn’t so supportive.
3. Did you have any contact with the program officer before submitting your application?
   I spoke with Hadley Wikham several times. The previous year, he was involved with the program so his insight was helpful as I prepared my application.
4. How did you learn of this fellowship opportunity?
   Professor Allen brought the fellowship to my attention.
5. What advice do you have for people thinking about applying for a graduate/predoctoral fellowship?
   Start early. In my case, writing is a process that takes many iterations and lots of time. Also, identify more than the required number of recommendation writers. Sometimes people become unavailable for one reason or another.
6. Anything else you’d like to add?
   Go Stanford Football!

Dane Van Domelon is pursuing a PhD in biostatistics at Emory University and was awarded the NSF graduate fellowship in March 2013. With the guidance of his research advisor, Robert H. Lyles, he is working on likelihood-based methods for logistic regression when one or more predictors are measured with error. He previously studied bioengineering at the Milwaukee School of Engineering and completed a two-year research fellowship in epidemiology at NIH.

1. Basics of fellowship: where, when, what, with whom, and what type of fellowship (NSF, NIH F30, NIH, F31, …)
   I received a fellowship through the National Science Foundation’s Graduate Research Fellowship Program (GRFP) in March 2013. The award provides a $32,000 stipend for three years, as well as a $12,000 cost of education allowance for the university.
2. Please describe your approach to preparing and writing your application.
   I tried to learn as much as possible about the application materials and the review process. I read articles online from previous GRFP fellows, and I got some great information from message boards. Then I spent a good chunk of time reading statistical literature and writing the Proposed Research essay. As for the Previous Research Experience and Personal Statement essays, I started with a list of the activities I wanted to remember to include, and then I wrote and revised the essays over the course of a few weeks.
3. Did you work with a faculty member on your application? If so, how?
   I approached my academic advisor for guidance three months before the application deadline. I had a general idea for the research proposal, but I had to learn a lot more about the topic before I could write a proposal that would be viewed favorably by statisticians on the review panel. My advisor linked me up with two professors in the department who specialize in the research area, and they directed me to relevant statistical literature and helped me think through the problem. I really couldn’t have asked for more support from the biostatistics faculty here at Emory.
4. Did you have any contact with the program officer before submitting your application?
   No.
5. How did you learn of this fellowship opportunity?
   I met a couple of GRFP awardees during my research fellowship at NIH.
6. What advice do you have for people thinking about applying for a graduate/predoctoral fellowship?
   Provided you have solid college grades and some substantial research experience, I would highly recommend applying for the GRFP. It’s a meaty application (three two-page essays, transcripts, and three letters of recommendation), but the risk-reward is very favorable. This past year, there were approximately 13,000 applications and 2,000 awards, a 15% success rate. The majority of applicants have two or even three years of eligibility, which further increases the probability of obtaining a fellowship. So even if your first application is unsuccessful, you can use the reviewers’ feedback to improve your application for the following year.
7. Anything else you’d like to add?
   Students are generally eligible to apply for the GRFP the year before enrolling in a graduate program and during the first two years of a graduate program. The application deadline is typically in mid-November.

Amanda Mejia is a third-year PhD student in biostatistics with Martin Lindquist and Brian Caffo at The Johns Hopkins University. She was awarded the NSF graduate fellowship in 2010. Her research is on the analysis of high-dimensional imaging data, such as fMRI and MRI. At the time of being awarded the fellowship, she was a PhD student in industrial engineering/operations research at Georgia Tech and was just beginning research in health care operations. Her NSF fellowship made the transition from engineering to biostatistics possible because of the freedom to work with different faculty and even change institutions or fields of study.

1. Please describe your approach to preparing and writing your application.
   It was my top priority for several months. I lightened my research and course load to focus on essays and meeting with recommenders. Georgia Tech provided a fellowship advisor, who gave me copies of past successful essays and taught me two important things. First, the NSF reviewers will only spend a few minutes reading your essays, so they have to be clear and concise, especially in the opening statement. Second, NSF’s guidelines for scoring applications in terms of intellectual merit and broader impacts are easy to find online. I explicitly stated in a bold font how my application met each criterion.
2. Did you work with a faculty member on your application? If so, how?
   I worked primarily on my own, with the help of Georgia Tech’s fellowship advisor. My faculty advisor read my research proposal, but only after I had nearly finalized it. It’s important to the NSF-GRFP that the applicant’s research proposal is their own idea, rather than their advisor’s.
3. Did you have any contact with the program officer before submitting your application?
   No, I didn’t have any contact with an NSF program officer beforehand.
4. How did you learn of this fellowship opportunity?
   Georgia Tech holds an information session for prestigious fellowship opportunities and another for the NSF-GRFP, specifically. I also knew some older graduate students who had the NSF fellowship, and their positive experience encouraged me to apply.
5. What advice do you have for people thinking about applying for a graduate/predoctoral fellowship?
   Make it a top priority, because it can make a difference in your career. Find someone who knows “the rules”—an advisor who has helped other students successfully apply or an older student who has gone through the application process.
6. Anything else you’d like to add?
   It’s hard to overstate the opportunities a prestigious fellowship like the NSF-GRFP can open up. My fellowship has allowed me to transition from engineering to biostatistics; it’s allowed me to work with different faculty and discover what research topic I want to pursue before committing; it’s allowed me to work for a summer in Munich without worrying about funding. I was even given the opportunity to go to Norway for a year through an international collaborative research program, which NSF is actively expanding. Applying takes effort and sacrifice, but it’s well worth it.

Jessica Trail is a doctoral candidate in the department of statistics at The Pennsylvania State University. Her research, under the supervision of Linda M. Collins and Runze Li, focuses on integrating techniques from functional data analysis and dynamical systems modeling to analyze intensive longitudinal substance use data. She received an F31 from the National Institute on Drug Abuse (NIDA) in 2012 and is funded for two years.

1. Please describe your approach to preparing and writing your application.
   After discussing with my advisors the possibility of writing a grant proposal and deciding to apply to NIDA for the F31 fellowship, we began by developing the project’s goals, which became the Specific Aims. The Specific Aims is an important piece of the application and influences the other sections of the proposal. To strengthen the application, I contacted a program official for feedback on my Specific Aims and reviewed a few successful F31 proposals from fellow Penn State students.
2. Did you work with a faculty member on your application? If so, how?
   I worked closely with my advisors, Linda Collins and Runze Li, who provided valuable feedback on my proposal and motivated and supported me throughout the process. Drs. Collins and Li collaborate frequently as principal investigators at the Penn State Methodology Center, an interdisciplinary research center that develops and disseminates new methods for research focusing on public health issues. Dr. Collins is studying the use of dynamical system approaches in behavioral intervention science. Dr. Li’s research uses functional data analysis approaches for analysis of intensive longitudinal behavioral data. My project was inspired by these two NIH Roadmap grants.
3. Did you have any contact with the program officer before submitting your application?
   I contacted a program officer from NIDA approximately six weeks prior to the submission deadline. She provided feedback about my Specific Aims and some general suggestions about how to make my application as strong as possible. I strongly recommend contacting a program official prior to submitting an application.
4. How did you learn of this fellowship opportunity?
   I approached my advisor, Dr. Collins, about the possibility of writing a grant proposal, or assisting in the writing of a proposal, during my research assistantship at The Methodology Center. Dr. Collins proposed the NIH Predoctoral Individual National Research Service Award, which provides predoctoral individuals with supervised research training in health-related areas.
5. What advice do you have for people thinking about applying for a graduate/predoctoral fellowship?
   Discuss available funding opportunities and your research interests with an advisor or other faculty member. They most likely know the process well and can provide much-needed support. Contact a program officer prior to submitting. It will hopefully generate interest in your research, and their feedback on your application is invaluable. Start early. It can take several months for the review process, and it may be necessary to revise your proposal and resubmit before the grant is awarded.
6. Anything else you’d like to add?
   I would like to thank the National Institutes of Health and the National Institute on Drug Abuse for the financial support of my research. I would also like to thank the Penn State Methodology Center for material and intellectual support of my research.