A recent Republican initiative asking citizens to review grants funded by the National Science Foundation gets the basic facts about several of those grants wrong, LiveScience has found.
In a video on YouTube, Rep. Adrian Smith (R-Neb.) calls for Americans to search the NSF database and report "wasteful" grants, citing two such projects: One, a $750,000 NSF grant "to develop computer models to analyze the on-field contribution of soccer players." The second "questionable" grant is one in which scientists "received $1.2 million to model the sound of objects breaking for use by the video game and movie industries." Smith's video is part of a larger Republican initiative called "YouCut."
But the researchers behind these projects say Smith has misrepresented their work and the amount of money spent on the projects.
"This was not $750,000 given by NSF for us to develop an algorithm to look at the performance of soccer players," Northwestern University engineering professor Luis Amaral told LiveScience. Amaral, who was the lead investigator on the soccer study cited by Smith, called the congressman's portrayal of his work "not only incorrect, but misleading."
"This was $750,000 that was given to a larger team of researchers to study a very broad range of questions related to creating provocative, efficient teams of researchers who innovate," Amaral said.
Cornell University computer scientist Doug James, the lead researcher of the sound-modeling study, had a similar reaction to Smith's characterization of his work.
"It is a gross misrepresentation of our activities and their intent," James wrote in an e-mail to LiveScience.
Soccer rankings and research dollars
Amaral's soccer study, published in June in the open-access journal PLoS ONE, was supported by two NSF grants. The first was a $450,000 award to develop efficient methods to evaluate the productivity of researchers and research institutions. The second was a $300,000 grant to study how teams collaborate. By quantifying researchers' contributions to their fields, Amaral and his colleagues hope to help funding agencies like the NSF allocate money more effectively.
How do those grants translate to studying soccer? According to Amaral, an M.D./Ph.D. student was rotating through Amaral's lab to learn the computer software Amaral and his colleagues use to model complex systems such as to explore how creativity and innovation arise from networks of researchers. The researchers decided to train the young scientist using easily available data from the World Cup. Soccer was particularly appealing, because team performance is difficult to rank using regular statistical methods, Amaral said.
"The soccer would enable us to look at teamwork," Amaral said. "Maybe we could learn from that how to understand how teamwork works. ... We just knew who was in a team and how the outcome was, but we didn't know how the team members had interacted."
In addition to the educational benefit of teaching the student the software and the possible wider implications of the work, the resulting rankings piqued interest in the private sector, Amaral said. Several European countries contacted him to ask about commercializing the technique.
"So it could be that this apparently useless, I don't know how they put it, 'wasteful,' research may even be resulting in the short-term in the creation of jobs and export of services and technologies to Europe," Amaral said.
Smith also got the price tag for the soccer research wrong, Amaral said. All three researchers on the paper were supported for their time by non-federal funding and fellowships, so little NSF money went toward the soccer ranking, he said.
"If I'm going to put a numerical value to what was actually spent on this particular research, these are very, very, very small amounts," Amaral said. The rest of the grant money is spent on his overarching line of research into teamwork and innovation.
Smith's second target, research to model the sound of breaking objects, is supported by an ongoing $1.2 million grant given to three researchers over four years. The goal of the research is to create advanced simulation technology for virtual environments, Cornell's James told LiveScience.
Right now, computers can't render sound the way they do graphics — sounds get dubbed in later. That makes it tough to match the sound to the action, making the virtual environment less immersive.
"I want to be able to simulate realistic virtual physical systems that look, move and sound realistic," James said.
The project on breaking objects was done over several months by a graduate student, James said, and was presented at the prestigious ACM SIGGRAPH conference in Los Angeles in July. As for the long-term applications of the work, James said, "the sky's the limit."
"Just think of the impact of computer-graphics rendering, and now imagine the combined potential for realistic computer-sound rendering," he said, citing possible uses of realistic simulations for engineering cars, aircraft and even spacecraft. The results may also be useful in designing rehabilitation and training simulations like those used in the military. Even robots could become better at navigating their environments with higher-level sound processing, James said.
James also defended research benefiting the movie and video game industry, a "nearly one-hundred billion dollar industry when combined," he said. Computer science students educated by NSF grants go on to innovate at influential companies like Pixar, he said.
"Attacking early research efforts as wasteful seems akin to telling 8-year-olds they shouldn’t waste their time dreaming about their future," James said.
Rep. Adrian Smith's office responded to LiveScience's queries with links to two press releases describing Amaral's and James' studies. When asked for clarification of how the research was singled out, Smith's spokesman, Charles Isom, declined further comment. Isom also declined to comment on LiveScience's findings about the research's cost and impact.
"Congressman Smith has always felt transparency is vital for the American public, and this video reflects his dedication to give taxpayers the opportunity to judge how their hard-earned money is being spent," Isom wrote in an e-mail to LiveScience.
Transparency in funding
The NSF requested budget for 2011 is $7.424 billion, an 8-percent increase over the previous year. According to the American Association for the Advancement of Science, federal science funding has remained flat for most of the last decade, even as the overall federal budget climbs.
In 2010, the total general budget was $3.55 trillion. In March, the Congressional Budget Office projected a 2011 deficit at $1.3 trillion. The NSF's $7 billion budget would equal approximately half a percent of that deficit.
The process of winning a grant through the NSF is "very rigorous," said NSF spokeswoman Maria Zacharias. Each year, the agency receives over 45,000 competitive grant proposals, and funds about 11,500, she said.
"Each proposal is evaluated in terms of not only its intellectual merit, but also its broader impacts," Zacharias told LiveScience. "The panels that are convened for peer-review are experts in the area of the proposal."
Both Amaral and James said they welcome more transparency — and public interest — in their work.
"It would be great if people actually read NSF project summaries, since there are some really fascinating things in there!" James said. "On the other hand, as implemented, it seems more like a politically motivated, anti-science witch hunt."
"It needs to be done in a sensible way," Amaral said, adding that he hoped the YouCut project would encourage people to learn more about academic research instead of jumping to conclusions.
"If I can be perfectly honest," Amaral said, "it is a case in which had the congressman sort of done his homework and tried to learn about the situation in a little more depth instead of just making kind of inflammatory statements, maybe it would have been better."
This article was reprinted with permission from LiveScience.
Related on LiveScience: