Dressed in satin and sequins, Roald Hoffman has ridden atop the first science-themed float in Rio's famed Carnival. Once a month, he appears on stage at a New York City café to host a revue of science and the arts. It's all part of what Hoffmann, a 1981 Nobelist for his work on the theory of chemical reactions, calls his “extreme outreach to the community.”

“I think science should be fun,” Hoffmann said in May to the National Science Board, the oversight body for the National Science Foundation (NSF), when it awarded him its prestigious Public Service Medal. But after flashing pictures of himself at Carnival and on stage at the Cornelia Street Café in Greenwich Village, Hoffmann got down to business: “Now I want to shift gears and talk about something serious.”

What Hoffmann wanted to discuss is a proposal for changing how the U.S. government supports the training of graduate students in the sciences. Federal research agencies now funnel most of their money for graduate students through grants to faculty members. That's the case for nearly 90% of the 39,000 graduate students whom NSF supports each year and for about two-thirds of those getting money from the National Institutes of Health (NIH). The remaining students are funded via fellowships, awarded directly to them, or through traineeships, in which universities compete for a grant to support a certain number of students in a particular area for a fixed period of time.

For the full text of this article, see the Science magazine Web site.

Jeffrey Mervis is a deputy news editor for Science magazine in Washington, D.C.

Jeffrey Mervis is a senior correspondent for Science magazine.

10.1126/science.caredit.a0900096