In 1999, when Robert T. Croyle, PhD and his colleagues at the National Cancer Institute (NCI) proposed funding transdisciplinary collaboration among tobacco researchers, the board of scientific advisers balked, said Croyle, speaking at an October conference on transdisciplinary science.

"They objected on the grounds that 'transdisciplinary' was a word we had just made up," said Croyle, now director of NCI's Division of Cancer Control and Population Sciences.

Times have changed, said Dan Stokols, PhD, a professor of Social Ecology at the University of California, Irvine. Today, transdisciplinary teams are earning large grants from private and federal agencies, he said. Fueling the trend, the National Institutes of Health (NIH) created new funding mechanisms for interdisciplinary research as part of the Roadmap Initiatives, launched in 2003. What's more, NCI and other institutes are increasingly collaboration friendly-allowing, for instance, multiple researchers to serve as co-principal investigators on grants.

However, little is known about what ingredients it takes to produce a successful cross-discipline collaboration, or even how to measure transdisciplinary teams' success. To address these knowledge gaps, organizations including NCI and APA's Science Directorate sponsored an October conference in Bethesda, Md., "The Science of Team Science: Assessing the Value of Transdisciplinary Research."

Behavioral scientists have much to contribute to improving team science, noted conference presenter Steve Breckler, PhD, APA's executive director for science.

"People are involved in transdisciplinary science, so it's not surprising that we need to understand the behavior of people," he said.

Forging new disciplines

There is a mounting need to evaluate the success of transdisciplinary teams, in part because they are so expensive, said Stokols. Productive transdisciplinary teams with ambitious aims require around $5 million over the course of five years, estimates William Trochim, PhD,a professor of human ecology at Cornell University. Therefore, Congress and private funding agencies want clear measures of success.

Among the most important indicators of success is rich team communication, said Karen Emmons, PhD, a clinical psychologist and professor of society, human development and health at Harvard University. Team members should spend time hashing out theoretical differences and even having arguments, she said.

"A good sign of interdisciplinary collaboration is discomfort," said Emmons.

That was the case with the Harvard Center for Cancer Prevention's project to identify and quantify cancer risk factors and then communicate their findings to the public. Epidemiologists, psychologists and risk-management experts butted heads over how to encourage people to change their behavior.

Their debates resulted in a novel solution-a Web site (www.yourdiseaserisk.harvard.edu) that calculates your risk and then lets you see it drop as you "try out" lifestyle changes, such as quitting smoking or eating more green vegetables, Emmons said.

The process often leads to a new, shared framework that transcends each of the individual's disciplines, Stokols noted. But, he added, that framework's just the beginning.

"Are these just novel integrations, or are they having an impact?" he asked.

How you measure that impact depends on the goals of the team. Some teams may aim to improve public health, by decreasing tobacco use, for instance. Others may be developing new conceptual models for complex behavior, integrating factors including public policy, individual behavior and genetic influences. Some transdisciplinary groups may even have the goal of developing new products, added Julie Thompson Klein, PhD, a professor of humanities at Wayne State University in Detroit.

"There is not a single form of goal, hence not a single [evaluation] criterion," Klein said.

Ingredients for success

An important aim of evaluating transdisciplinary teams is developing a knowledge base of what it takes for such groups to flourish. As new teams form, scientists may someday be equipped to perform a "collaborative readiness audit," Stokols said.

So far, research suggests that successful transdisciplinary teams have:

  • Institutional support. Funding mechanisms that encourage cross-disciplinary collaboration lay the foundation for such research, but researchers' home universities also play an important role, said Emmons. Tenure decisions should take into account how well scientists collaborate with others and whether they make contributions outside of their own field, she said. Universities can also support transdisciplinary groups by providing space for local groups and cyber-infrastructure for far-flung teams, Stokols said.

  • Visionary leaders. One of the biggest challenges of transdisciplinary groups is keeping people with very different backgrounds on the same page, said Barbara Gray, PhD, an organizational behavior professor at Penn State University's Smeal College of Business.

"Leadership needs to knit together individual vantage points into a new coherent whole," she said.

Such leaders need to be fluent in different research languages and help translate what, for instance, a microbiologist is saying in terms that make sense to an epidemiologist.

  • Collaborative members. Leaders can't do all the work of bridging disciplines, noted Greg Farber, PhD, a health scientist administrator at NIH. Training programs for graduate students should prepare them to work with people outside their fields. Social scientists, for instance, might take classes on the math of computer simulation, he suggested.

Increasing knowledge about what fosters productive transdisciplinary collaborations and then putting that information to use is a daunting proposition, but it's crucial to continued advancements in science and public health, Farber said.

"In the field of biomedical science, we have already collected many of the low hanging fruit," Farber said. "We really need to do more to foster that vertical integration."