Broader Impacts - Should You Care?

Expanding the Pipeline

By Tracy Camp and Juan Gilbert

Yes! For many reasons, you should definitely care about broader impacts.

First, many CISE researchers report that broader impact efforts bring inspiration, personal satisfaction and new perspectives on their work. What could be more rewarding than seeing significant impact from your efforts? Second, if you receive federal funds for your research, then you should feel a moral obligation to return the taxpayers' investment by participating in efforts that will ultimately benefit society. Taxpayers deserve benefits for the money that they provide. Third, on May 28, 2010, the America COMPETES Reauthorization Act of 2010 bill was passed by the House and is currently being considered by the Senate. If passed, NSF will need to implement a policy for the Broader Impacts Review Criterion that meets the goals of the Act. (For further details on COMPETES, see: "NSF’s Broader Impact Criterion," Computing Research News, vol. 22, No. 4, p. 6, September 2010.)

Several leaders within CISE want (and, potentially, need) current and future investigators to improve the broader impacts of each funded NSF grant. While the intellectual merit contributions from CISE investigators appear to be strong, the broader impacts could be (and should be) improved. The five broader impacts criteria are:

1. advance science while promoting teaching, training and learning;
2. broaden participation of underrepresented groups;
3. enhance infrastructure for research and education;
4. provide broad dissemination to enhance scientific and technological understanding; and
5. highlight the benefit to society.

There was, however, much confusion among members of the CISE community about what these criteria really mean. Therefore a two-day Summit was organized in D.C. beginning the evening of June 21 to discuss, present, and subsequently develop guidance materials for the NSF computing research community on how to effectively integrate broader impact activities into research projects. Funding was provided by the Computer and Network Systems (CNS) Division of CISE. Attendance was by invitation only. Approximately half of the 123 attendees were invited via a submission process. Attendees included CISE researchers, NSF Program Managers, NSF Division Directors, Ph.D. students (who helped document the working group sessions), and Computing Innovation (CI) Fellows (who have written report drafts that summarize the discussions of each broader impact category).

Keynote speakers were Jeannette Wing (then Assistant Director, CISE, NSF), Jim Shelton, (Assistant Deputy Secretary for Innovation and Improvement at the Department of Education), Neil Gershenfeld (Director of the Center for Bits and Atoms at MIT), and Deborah Estrin (Director of the Center for Embedded Networked Sensing at UCLA). In addition to the keynote speakers, 24 other presentations were given by CISE community members, each of whom was asked to showcase a particularly impressive broader impact activity.

All materials developed for the Summit are posted at: http://cisebroaderimpacts.org/. Currently this CISE Broader Impacts site includes:

1. The full Summit agenda, with videos and slides of all Summit presentations.
2. A list of the people who participated in the Summit. We encourage you to contact anyone on this list and request assistance in improving the broader impacts of your work.
3. Examples of high-quality broader impact activities. We detail one activity from each of the five broader impact categories, as well as one activity that spans multiple categories. We also provide links to other examples of activities that were presented at the Summit.

One example of a high-quality broader impact activity from category 4 (i.e., provide broad dissemination to enhance scientific and technological understanding) concerns giving presentations to the public via science cafes. Miriah Meyer (CI Fellow at Harvard University) writes:

“For the cost of a cup of coffee or a pint of beer, people interested in various scientific topics converge to hear talks by local scientists and engineers. These informal talks are a chance for researchers to bring their ideas and work to the local community. Examples include domestic grassroots events announced on the Nova ScienceNow webpage, and the Cafe Scientifique events in the UK sponsored by the Wellcome Trust. Many of these events are founded by university groups, such as the Boulder Colorado Cafe Scientifique.”

In the near future, the CISE Broader Impacts site will also include:

1. Reports that summarize the presentations and discussions at the Summit for each of the five broader impact categories. The main goal of these reports is to document both existing superior broader impact activities and new innovative ideas for broader impact activities.
2. A document that offers suggestions on how NSF proposal reviewers should evaluate proposed efforts involving broader impacts.
3. Opportunities to forge collaborations and partnerships with the potential to enhance the broader impacts of computing funded research projects. 

We strongly encourage the CISE community to begin putting more emphasis on broader impacts in both their proposals and funded projects. In fact, we expect that it is very likely the NSF will be placing a greater emphasis on broader impacts in the proposal review process. While the broader impacts are clear with certain types of NSF-funded grants (e.g., education-related grants), technical research grants need broader impacts as well. The main goal of this effort is to provide the computing community with examples and materials on ways that computer scientists can have broader impact on their research, education, and wider communities. It is extremely important that we find ways to demonstrate the broader impacts of our work, as some argue that science is at a crossroads and in some instances being attacked. What better way to justify science than through broader impacts? We invite you to get involved.

Tracy Camp is Professor and Interim Department Head of Mathematics and Computer Sciences at the Colorado School of Mines.
Juan Gilbert is a Professor and Chair of the Human-Centered Computing Division in the School of Computing at Clemson University.