AAAS Big Data Policy Fellowships: Great Opportunity for Statistical Scientists to Shape Federal Big Data Policy

By Steve Pierson posted 08-28-2013 12:24

[8/19/14 update: Applications for 2015-2016 Executive, Congressional and Judicial Branch Fellowships are due November 1: To see the 2014-15 Big Data and Analytics Fellows, click here.]

If you've ever felt statistics needed a larger voice in federal policy, the AAAS Science & Technology Policy Fellowships may be your opportunity! This highly successful and influential AAAS program, which places PhD scientists in Federal agencies or Congressional offices for a year, started forty years ago but new for 2014-2015 is a Big Data & Analytics track as one of their nine Fellowship Program Areas. I highly encourage statistical scientists to apply for this fellowship, for which applications are due November 1.

Before I discuss further the AAAS S&T Policy Fellowship for Big Data & Analytics, let me direct you to a 12/17/12 blog entry in which I urge statisticians to apply for science policy fellowships: Science Policy Fellowships: Statisticians Should Apply. There you will find links to other fellowship opportunities, in addition to the AAAS one. I also just became aware of the White House Office of Science and Technology Policy Internships.

For health-related fellowships, statistical scientists should look at the Robert Wood Johnson Foundation Health Policy Fellows program, which is described as an "opportunity for exceptional mid-career health professionals and behavioral and social scientists with an interest in health and health care policy. Fellows experience and participate in the policy process at the federal level and use that leadership experience to improve health, health care, and health policy. The fellowship is a residential experience in Washington, DC, with additional support for continued health policy leadership development activities." The call for applicants will go out September 6 with applications due November 13.

For the Big Data and Analytics fellowship, AAAS expects 5-15 placements in any federal agency in Washington, D.C. that partners with AAAS to "address a broad range of policymaking and implementation challenges by applying your analytical skills to data and trend analysis issues from infrastructure, technology, quality control, and presentation to security, integrity, and ethics." 
In an August 14 AAASLive Event, AAAS invited Dr. Carolyn Lauzon, a current AAAS S&T Policy Fellow at the Department of Energy in the Office of Advanced Scientific Computing Research (ASCR), and Dr. Mark Peterson, a former AAAS S&T fellow and who currently leads the Data and Analytics team within USAID’s Office of Science and Technology, to join the program.

In the extensive discussion, I found three questions that could be useful to statistical scientists interested in the Big Data & Analytics fellowship. Please see the full discussion for more information.

Peterson replied to a question on what projects a Big Data fellow will be working on this way:
I expect this to vary quite a bit depending on the background of the individual fellow and the needs of the office where she's placed. On my team at USAID, we're looking for fellows who will be able to take the lead in answering questions important for making decisions and guiding strategies in our overseas missions and our Washington-based offices as well. As one example, a current fellow is working on a project to measure deforestation in the Andean Amazon region. This is part of a multi-country, multi-year program with many partners. So a key part of her work is to work with those partners to bring together many disparate data sets in a platform that allows all of the partners to interact with those data sets. Another part is using those data sets to answer questions about how well we're doing in reducing deforestation, identifying the most effective interventions, etc.
Lauzon added:
First, I’m in the DoE Office of Science which funds basic science research and manages 10 of our nation’s national labs, so when talking about 'Big Data' my office is talking about extreme scale science. For example, data coming from our light sources is entering the terabyte/hour range and that rate is only expected to grow. One important challenge is that this incredible data growth (rate of data being produced and the data volume) makes data management and data analysis increasingly difficult. Yet data management is crucial for the scientific method. Reproducibility, transparency, open-access for peer evaluation, and continued knowledge discovery from existing data all depend on good data management.

To address these challenges I've worked with the Office of Science to find methods to encourage scientists to spend the time and resources necessary to appropriately manage data. We also dialogue with members at our national labs about defining their role in this era of ‘Big Data’. I help out with this group doing mini research projects and engaging with contractors at our facilities. I also work with my specific office, Advanced Scientific Computing Research (ASCR) which manages supercomputers and a high bandwidth scientific network at our facilitites. If you have extreme scale data you will want extreme scale computational resources and ASCR has a role to play here. ASCR also makes research investments such as advanced methods for data analysis (e.g. advanced visualization techniques). So I work with ASCR to learn the needs of the other offices and understand how to target those needs, and brainstorm on future infrastructure for managing big data.
To a question about what skills will make an applicant competitive for the Big Data & Analytics fellowship, Peterson responded
The skill set needed will depend on the position the fellow ultimately is placed into. I think the most important qualifications will be a deep understanding of issues around using (generating, sharing, storing, analyzing, communicating about) data and the ability to manage a project focused on data management and analysis to successful completion. That may or may not require advanced computer skills (though that would never hurt), depending on how deep the fellow needs to get into technical details of either data analysis or the design of platforms and tools for working with data.
To this question, "Relational database development as well as statistical modeling in general, can be applied to all areas of science. Would you suggest identifying a specific area of interest (e.g., education, energy) in the application, or a general interest in data analytics?", Peterson wrote,
I think that if you can communicate an understanding of how your experience and interest in data analytics would translate to work in specific sectors (e.g. education, energy), that can help prospective offices understand where you might fit. Some offices might be looking for a very specific set of skills (e.g. relational database development) or experiences (e.g. in computational genomics), but most will not expect to see a perfect match among the candidates and instead will be looking for someone who has enough of the technical background paired with the ability to quickly learn how to navigate a new area. Enthusiasm for a particular office, agency, or sector could be part of what makes a candidate seem like a good match, but I think it's more important to highlight the skills and experience you bring and your ability / interest in taking on new challenges. (Of course, if you are aiming for a particular area or type of work, no harm in communicating that.)
See also these previous blog entries relating to big data: You might also be interested in this 1/14/13 blog entry based on a piece by a former AAAS Congressional Science Fellow: Advice for Dealing with Congress from Applied Mathematician Who Spent Year on Capitol Hill.
See other ASA Science Policy blog entries. For ASA science policy updates, follow @ASA_SciPol on Twitter.