The Science of Team Science

Feature The Science of Team Science

BETH BAKER

An emerging field delves into the complexities of effective collaboration.

s research questions grow ever collaborate with others who have com- team science, as measured by publi- Amore complex and science plementary expertise, in the context of cations (eliminating self-citations as struggles to swim through big data, science now. We now have access to a factor) that are more highly cited major funders, including the National more data, and we think about major compared with single-author papers, Science Foundation (NSF) and the scientific and social challenges more more patents, and the production of National Institutes of Health (NIH), holistically—with a systems approach,” more novel work. are pushing scientists to collaborate says psychologist Kara L. Hall, with the David Foster, director of Harvard more across disciplines, institutions, National Cancer Institute (NCI) at the Forest, points to a study published in and even nations under the banner of NIH and a pioneer in SciTS. May in Nature Climate Change as an team science. The scientific advances In the life sciences, major research example of what a large cross-disciplin- made through cross-disciplinary initiatives—from completing the ary research effort can produce. Part research are crucial. But researchers Human Genome Project to identifying of the Long Term Ecological Research are finding that doing team science the hantavirus or understanding the (LTER) and Science Policy Exchange well can be challenging. The new field effects of climate change—have been programs, the new study looked at three of the science of team science (SciTS) tackled by researchers collaborating scenarios for power-plant emissions and seeks to shed light on what makes across disciplines and institutions. A analyzed the resulting public health ben- effective teams in order to produce the newly released report by the National efits, as measured by the number of pre- best outcomes. Research Council (NRC), Enhancing mature pollution-related deaths. Such “There are so many reasons why the Effectiveness of Team Science, cites research could not be done by those in a a given individual really needs to studies demonstrating the value of single discipline, Foster suggests.

In 1993, a transdisciplinary research effort by medical experts, mammalogists, evolutionary biologists, and climatologists led to the rapid identification of the deadly hantavirus, found to be carried by the deer mouse. Photographs: Cynthia Goldsmith, Centers for Disease Control and Prevention (hantavirus); Seney Natural History Association (deer mouse).

Downloaded fromhttp://bioscience.oxfordjournals.org https://academic.oup.com/bioscience/article-abstract/65/7/639/258550 July 2015 / Vol. 65 No. 7 • BioScience 639 by Wilson Dental Library user on 22 June 2018 Feature

“The insights are so overwhelm- ingly positive,” he says of his par- ticipation in the LTER program and other cross-disciplinary teams. “For example, I’m an ecologist working in New England, and like every other landscape, it’s been shaped by physical, cultural, and biological processes. I do the ecology, and there’s nothing like having a great physical scientist who understands climate systems and archaeologists who understand the role of people in the environment. The levels of insights are fantastic.” The team science train has clearly left the station, as the NRC report shows. In 1959, a bare majority of papers had more than on author. By 2013, 90 percent did. A 2010 study in Science of Translational Medicine looked at more than 21 million papers published from 1945 onward and found “a fundamental and nearly universal shift in all branches of science: Teams increasingly dominate solo scientists in the production of high-impact, highly cited science.” Moreover, teams are growing in size and in number of institutions rep- resented, with a majority written by 6 to 10 individuals from more than one institution. A 2007 study by Stefan Wuchty and colleagues, published in Science, which looked at nearly 20 million research articles across academia and 2 million patents, found that teamwork was greatest in the life and physical sciences. But measuring the overall productivity and outcomes of team science is tricky. For one thing, says Jonathon Cummings, an organizational researcher in the business school at Research mentor Paul Siqueira works with a team of Harvard Forest summer Duke University, who served on the research students to calibrate a motorized tramway with sensors that NRC SciTS committee, “what we see will measure light, humidity, moisture, and other attributes above is the patent or the publication, and a regrowing forest. Research mentor Andrew Richardson prepares to head we do not observe all the failed col- into the tree canopy to sample vegetation with students Sidni Frederick laborations. When you just look at the and Ivonne Trujillo to help validate and scale remote measurements of tree observed artifacts, there is a relation- phenology. Photographs: Grace Barber, courtesy of the ship between having more authors and Harvard Forest. having higher impact, but it’s hard to know if it’s causation versus correla- better evaluative tools to measure team processes and products, and surveys tion. So there could be a bias in the effectiveness “that includes, but is not and interviews of team science partici- s amp l e .” limited to, bibliometric indices, co- pants. Particularly challenging is the The NRC committee stressed the authorship network analyses, experts’ measurement of deep interdisciplinary need for an effort to develop more and subjective appraisals of team science knowledge integration.”

Downloaded from640 https://academic.oup.com/bioscience/article-abstract/65/7/639/258550 BioScience • July 2015 / Vol. 65 No. 7 http://bioscience.oxfordjournals.org by Wilson Dental Library user on 22 June 2018 Feature

The biggest challenge, says Hall, of the field, team members need to is interpreting the value of scien- be able to clearly explain their point tific research—in terms of not just of view, ask basic questions to clarify the number of citations but also the concepts, and practice active listening. effect that the research has on solv- Team science is different in some ing scientific or real-world problems. important ways, though, especially in For organizations such as the NIH, the independence researchers enjoy. she says, evaluators ultimately want to An academic science researcher has know, “Are we enhancing the health of the freedom to assemble a team and the nation? How we link the products the freedom to refuse to participate in of science teams to real-world health another researcher’s project. impacts is a significant challenge.” “In other organizations, the leader Beyond the effort to measure the has the final say,” explains Fiore. “In outcomes of team research, SciTS hopes a business, you can fire somebody. to guide scientists in building effective In universities, you have tenure, and teams. How does a researcher decide that alters the organizational dynam- whether a particular problem is best ics. Also, in universities, there’s a dif- studied by a team or by an individual? ferent influence and power structure.” If a team is best, should members cross Department chairs, for example, may or disciplines and institutions? How large may not play a role in leading science. should it be? Who will lead it? Is there “You may have department chairs who Cognitive scientist Steve Fiore enough grant money to pay for extra are not as active in the field as others stresses that it is crucial for teams time spent on communication and within that department. You could have to invest time in developing a coordination, travel to in-person meet- a professor in a biology department shared understanding and language ings, or special team training? And will who is world famous and [the] chair in order to avoid conflict and team researchers receive proper credit, could be less well known, and that miscommunication. either in terms of promotion and ten- creates an unusual dynamic… with dif- Photograph: Allison W. Fiore. ure or publication authorship? ferences like this, you can’t simply take theories from the study of leadership in The creation of a new field other domains and directly map them does not fit all…. It’s a nuanced and Although there have always been teams onto science.” complex set of activities.” of scientific collaborators, the idea of How firm is the SciTS research studying specifically what makes a foundation? SciTS experts interviewed Are there benefits to diversity? scientific team effective took off just for this article, most of whom come Much of SciTS focuses on how to a decade ago. The term SciTS was from the social sciences, range in their work across disciplines, institutions, or coined in 2006 at a conference, The opinions, from viewing the science as geography. The greater the diversity of Science of Team Science: Assessing the in its infancy or early adolescence to the team, the greater the challenges— Value of Transdisciplinary Research, seeing it as “emergent” or well estab- but also, if done well, the greater the hosted by the NCI. In 2010, the first lished. “This is a field really strongly impact. “Diversity brings about new annual international SciTS conference empirically grounded,” maintains ideas and challenges assumptions we was convened. In June, the sixth such microbiologist Holly Falk-Krzesinski, make. If we have a shared framework conference was to be held at the NIH. who began working on SciTS while at and language, that’s very helpful in SciTS draws heavily on team Northwestern University and who now some cases and makes for an enjoy- research from other fields—business, is a vice president at Elsevier. “From able team experience,” says Noshir the military, sports. “One of the first my perspective, versus other kinds of Contractor, professor of behavioral questions we are trying to address in endeavors of science, this has such a sciences at Northwestern University. SciTS is what do we already know strong foundation of scientific evi- “On the other hand, a team with [about collaboration] and how it can dence.… Teams provide a really good diverse backgrounds can help chal- help us understand and improve scien- opportunity to be highly impactful.” lenge you and make you explain what tific teamwork,” says Stephen M. Fiore, Still, says AIBS board member and you mean.... That is another reason with the University of Central Florida’s ecologist James P. Collins of Arizona that diversity can help. It’s also very cognitive science program, who, along State University, the science of team risky. Sometimes it crashes and burns.” with Hall, served on the NRC’s SciTS science is complicated. “It would be In a 2013 study published in committee. For example, “We already a mistake to try to characterize team Psychological Science, Cummings and know from decades of study the impor- science as ‘this thing,’ without appreci- colleagues found that bigger may not tance of communication.” Regardless ating the variance,” he says. “One size be better when it comes to scientific

Downloaded fromhttp://bioscience.oxfordjournals.org https://academic.oup.com/bioscience/article-abstract/65/7/639/258550 July 2015 / Vol. 65 No. 7 • BioScience 641 by Wilson Dental Library user on 22 June 2018 Feature

research. On one hand, larger science being successful,” says Cummings. groups, made up of members from “[Team science is] complementary and different disciplines and institutions, not mutually exclusive of individual on average are more productive than investigator–driven research,” agrees are smaller teams, “but their marginal Falk-Krzesinski. productivity declined as their hetero- Collins adds that many, if not most, geneity increased.” The researchers scientists conduct both individual and concluded that “group heterogeneity team research over the course of their moderates the effects of group size, careers. “It’s not either–or,” he says. and… desirable diversity in groups Some worry, though, that funders may be better leveraged in smaller, may be tipping the scales toward more cohesive units.” Cummings teams. “There is a continuing con- believes that the ideal number of team troversy about whether or not [team members is six to nine, because larger science] squeezes out the individual teams become more bureaucratic, even researcher,” says Julie Thompson within the same discipline. Klein, a professor of humanities at Citing an influential 2008 “Ecology Wayne State University and an expert of Team Science” article, published in interdisciplinarity. “You need teams in the American Journal of Preventive to solve complex problems, but what Medicine, the NRC report found that role does the individual researcher Holly Falk-Krzensiski, of Elsevier, “large multi-institutional team science play? I’ve talked to scientists at lesser- says that interdisciplinary teams are projects are highly labor intensive, endowed universities, and they get especially needed to address large, prone to conflict, and require substan- squeezed out of funding. Their work intractable problems. tial preparation and trust among team is not regarded at the same level. But Photograph: Courtesy of members to even partially achieve their the role of the individual scientist is Holly Falk-Krzensiski. scientific and translational objectives.” important.” Teams whose members are geo- Indeed, Cummings maintains that graphically dispersed face special the main driver of team science is Given these challenges, how does a challenges. The research shows that the funding community rather than researcher launching a project decide face-to-face meetings are important researchers eager to join arms. “There’s whether to approach it alone or with a for forging good working relationships so much money to be had, funding- team? “They don’t know if it’s worth it, and building trust. For international wise, that some researchers bear the and that’s the problem,” says Cummings. teams whose membership spans time burden of [team science] in order to “Part of it is the working relationship…. zones, even making appointments to get the money.” Then there’s the actual science—does Skype can be difficult, let alone finding it complement what you need, and do the time and money to meet in person. But is it worth the trouble? you have something they don’t have? I Contractor says it has been “shown A 2013 study published by Eric D. take a two-pronged view at the types of pretty consistently working with peo- Roy and colleagues in BioScience relationship: knowledge and expertise, ple at different locations gets higher (http://io.aibs.org/roy) surveyed natu- and [personal] relationships.” impact. However, it has also been ral and social scientists to learn what In some cases, an interdisciplinary shown that NSF projects that have obstacles prevent effective research team is mandated by a funding agency investigators who are geographically on environmental problems: “The in its request for proposal. In other dispersed systematically do worse. Is respondents identified many advan- cases, researchers may be motivated this a paradox? Not really.… Most tages and rewards of interdisciplin- to work with a team because such an of the time, geographic dispersion ary research, including the creation approach flows organically from the undermines the capacity for research. of more-relevant knowledge. However, problem itself. But when it does succeed, it succeeds they also reported significant chal- Plant ecologist Mark Schwartz, spectacularly.” lenges and obstacles, including tension who directs the John Muir Institute of with departments (49%) or institutions the Environment at the University of Solo or team? (61%), communication difficulties, and California, Davis, has worked on many Despite team science’s ascendency, no differing disciplinary approaches, as cross-disciplinary collaborations. “The one is suggesting that research by lone well as institutional barriers (e.g., a lack challenge is can you justify the pay- investigators is less valuable, SciTS of credit in promotion and tenure). off or the benefit of that interaction experts hasten to add. “There are great Most (52%) believed that developing against that cost of extra time? There examples of lone geniuses being suc- interdisciplinary breadth should begin are some cases where it’s a drop-dead cessful and great examples of teams as early as the undergraduate level.” easy answer: yes, it’s worth it,” he says.

Downloaded from642 https://academic.oup.com/bioscience/article-abstract/65/7/639/258550 BioScience • July 2015 / Vol. 65 No. 7 http://bioscience.oxfordjournals.org by Wilson Dental Library user on 22 June 2018 Feature

When a group of biologists was studying managed relocation of species, they began by suggesting various ecological models. “But it was still largely an academic exercise,” he says. “When it came down to it, it became important to ask, ‘What do the laws say? How do you put it in an ethical framework?’ You couldn’t do this issue without bring- ing in all these other disciplines. So it became a cross-disciplinary effort, and that went really well.” In contrast, he says, his work with the Southwest Regional Climate Hub, which is made up of government agen- cies, academic researchers, and agri- cultural interests, sometimes leaves him longing for a unidisciplinary approach: “The hub is inherently interdisciplinary, but there is a large focus on crops, and I just don’t have much to contribute to that conversa- Land management is a multidisciplinary team effort. Shown here (from left) tion. Then, I find myself wishing we are ecologist Virginia Boucher, lead staff person for the University of California, were talking about trees,” he says. Davis, Natural Reserves; chancellor Linda Katehi, an engineer; plant “For some research areas, [team sci- pathologist Neal Van Alfen, former dean; and Shane Waddell, reserve manager. ence] emerges naturally, but for a lot Boucher and Waddell are demonstrating the solar meteorological and wireless of others, I don’t think it dawns on network system used to track animals and climate. Photograph: Mark Schwartz. you,” says biologist Warren Burggren, with the Developmental Integrative Further reading. Biology Research Cluster at University of North Texas. He collaborated with Cooke NJ, Hilton ML, eds. 2015. Enhancing the Effectiveness of Team Science. National a mathematician on a paper on physi- Research Council, Committee on the Science of Team Science, National Academies ological complexity that came about Press. through an informal office friendship, Science of Team Science (SciTS), a collection of more than 1800 research papers, is a site batting around ideas on Friday after- initiated by Holly Falk-Krzesinski of Elsevier. www.mendeley.com/groups/3556001/ noons. “I would not have sat down to science-of-team-science-scits write this myself—it was much stron- Team Science Toolkit, by the National Cancer Institute, includes best practices and ger because of the involvement of this policies., www.teamsciencetoolkit.cancer.gov/Public/Home.aspx. applied mathematician,” he says.

Qualities of a well-functioning team Variations on a team. Personal relationships can be the key to team cohesion. Harvard Forest’s Unidisciplinary. Members are from a single discipline. Foster says, “The single most impor- Multidisciplinary. Team members from different disciplines make separate contributions. Interdisciplinary. Team members from different disciplines integrate “information, tant thing is you don’t select the data, techniques, tools, understanding.” best credentials; you select the best Transdiscplinary. The team integrates and transcends disciplinary approaches to person who will fit in,” he says, “and come up with new understandings. because they fit in, they thrive. It’s (Adapted from Cooke NJ, Hilton ML, eds. 2015. Enhancing the Effectiveness of Team Science, painful to learn that. But if you’re from the National Research Council, Committee on the Science of Team Science.) smart, you don’t have to learn it too many times.” One of the first challenges a new The need to clarify language “slows high-level goals and others meaning cross-disciplinary team faces is clarify- everything down,” says Schwartz. One near-term targets. ing vocabulary and goals. The most group with whom he worked became The NRC report found that when basic words—model, network—mean bogged down over the word objective, diverse groups—whether in terms different things to different disciplines. with some using the term to mean of discipline, employment sector, or

Downloaded fromhttp://bioscience.oxfordjournals.org https://academic.oup.com/bioscience/article-abstract/65/7/639/258550 July 2015 / Vol. 65 No. 7 • BioScience 643 by Wilson Dental Library user on 22 June 2018 Feature

geography—“fail to identify, discuss, lot of smoke but not a lot of flames in and clarify… differences among mem- universities,” says Burggren. bers, confusion and conflict can arise.” Tenure and promotion policies still Building in that extra start-up time favor the accomplishments of the indi- and becoming adept at team processes vidual researcher. “We don’t know very are fundamental to a team’s success. In well what the career impact is for the business world, these lessons were those working on these teams,” says learned decades ago, but scientists are Cummings. “Are the PhD students often poorly trained in collaboration. and postdocs and junior or senior “What we’re woefully lacking is training professors more likely to make con- and expertise that helps us be effective in tributions? From a human capital per- an interdisciplinary team,” says Schwartz. spective, what are they gaining from team science that they would not oth- A culture change in academia erwise gain from small-scale projects?” “A big challenge in our scientific Wayne State’s Klein has long called enterprise is that we have a culture of on universities to support mentoring, competition, where what we need is to provide physical space for collabora- more cooperation,” says Hall. SciTS tion, and to institute cross-disciplinary experts point to challenges in aca- training. “If the field is going to have any long-term impact, it needs to be demia—including a lack of organiza- Kara Hall of the National Cancer institutionalized, and that’s extremely tional support, departmental silos, and Institute helped launch the SciTS uneven,” she says. She worries that promotion and tenure policies that do field by organizing the first Science of smaller or poorer schools will be left not reward collaboration—as barriers Team Science Conference in 2006. off the team bandwagon. “The great to working in interdisciplinary teams. Photograph: Jeremy Rusnock. In the private sector, too, labs compete danger is the divide between the haves over patents or new drug development and the have-nots,” she says. and are reluctant to share data. Many point out the importance of training,” says Klein. “So it’s absolutely The NRC report highlights some teaching students early on to work in crucial now to look at how well or not bright spots. Many universities have cre- collaborative teams. (Hall says she is graduate education prepares students, ated new science teams, larger groups, pleased that her third-grade son gets future faculty, future lab workers—the and research centers. Arizona State, graded not only on reading and writ- whole array of people resources in for example, is cited as a national pace- ing but also on collaboration.) But little interdisciplinary work.” setter in restructuring the university, empirical research has been conducted investing in interdisciplinary schools to see whether cross-disciplinary pro- and research centers such as a school of grams aimed at preparing university biodesign and a school of sustainability. students for team science are even Beth Baker ([email protected]) is a freelance writer and the Features editor of BioScience. Still, academic science has a long effective, the NRC report notes. Her latest book is With a Little Help from Our way to go to forge a collaborative “You don’t really change things Friends: Creating Community as We Grow culture across disciplines. “There’s a if you don’t change education and Older.

Erratum Owing to an editing error, the Feature article by Myrna Watanabe about the Nagoya Protocol (BioScience 65: 543–550) misstated the ability of the United States to approve the protocol. Because the United States has not ratified or acceded to the Convention on Biological Diversity, it could not sign the protocol before it came into force and cannot now accede to it unless it first accedes to the Convention on Biological Diversity, which would require a two-thirds vote of the Senate and the president’s signature.

Downloaded from644 https://academic.oup.com/bioscience/article-abstract/65/7/639/258550 BioScience • July 2015 / Vol. 65 No. 7 http://bioscience.oxfordjournals.org by Wilson Dental Library user on 22 June 2018