In Chapter 3, the committee concluded that training interventions offer a promising route to increase team effectiveness. This chapter builds on that conclusion, reviewing research on team training and education for team science. The chapter begins with an introduction to team training, its goals and effectiveness. The second section reviews team-training interventions that show promise for increasing the effectiveness of science teams and larger groups, and the third section reviews interventions designed specifically for team science. The fourth section focuses on education for team science. The fifth section reviews training and education strategies that can help to address the challenges emerging from the seven features introduced in Chapter 1. The chapter ends with a summary, conclusions, and a recommendation.
As a preface to the chapter, we note that professional development, education, and training are general terms that are too often used without clear definitions. The terms “training” or “professional development” can be used to describe a variety of learning activities, ranging from an hour-long presentation on a given scientific topic to a weekend retreat about managing team conflict. The word “education” might be used to describe the same hour-long presentation on a scientific topic or an undergraduate course designed to teach students from different disciplines how to work together on team projects. The context can provide some clues. In universities, the terms “professional development” or “training” are typically used to describe activities outside the classroom, such as research experiences, while the word “education” refers to in-class learning experiences. But,
even in academic contexts, confusion can arise. For example, when doctoral students attend an hour-long presentation on a scientific topic related to their research, should the learning experience be called education, professional development, or training? When postdoctoral fellows, who have completed their formal education, attend the same presentation, should it now be called professional development or training?
In sum, the use of the terms “education” and “training” both in the research literature and in practice can sometimes be arbitrary, although which term is used may affect how learning processes and outcomes are measured and funding is allocated. Despite these important distinctions, for sake of reviewing the literature in this chapter, we use the terms adopted by the authors of each study. In future research, it will be important to delineate more clearly the meaning of these teams to develop greater coherence in science policy and practice.
Generally, team training is defined as an intervention to improve team performance by teaching competencies necessary for effective performance as a team (Cannon-Bowers et al., 1995; Delise, Gorman, and Brooks, 2010). Drawing from the decades-long tradition of learning research in psychology and education, Kraiger, Ford, and Salas (1993) argued for organizing the desired learning outcomes of training in terms of knowledge, skills, and attitudes. The same three categories of learning outcomes have been adopted in the team-training literature, as follows (Cannon-Bowers et al., 1995; for reviews, see Salas et al., 1999; Salas, Cooke, and Rosen, 2008; Klein et al., 2009; Delise, Gorman, and Brooks, 2010; Shuffler, DiazGranados, and Salas, 2011):
- team knowledge (e.g., task understanding, shared mental models, role knowledge)
- team skills (e.g., communication, assertiveness, situation assessment); and
- team attitudes (e.g., team orientation, trust, cohesion).1
Training for a particular team is often designed based on analysis of the situational and environmental context, which establishes team goals and tasks and enables identification of the needed knowledge, skills, and attitudes (Bowers, Jentsch, and Salas, 2000).
1 Research on educational preparation for team science has also organized the desired learning outcomes into these same three categories, as discussed later in this chapter (e.g., Nash, 2008).
Recent research provides a more detailed framework of team knowledge, skills, and attitudes (which we refer to as “competencies”) emerging from the team context, as well as the situational and environmental context that can be used to design training strategies. First, team training may focus on either taskwork or teamwork competencies (or both). Taskwork training targets the improvement of task-specific competencies (for science teams and groups, this would include scientific knowledge and skills related to the research problem), while teamwork training targets the improvement of team collaboration competencies. Building on the distinction between taskwork and teamwork proposed by Cannon-Bowers et al. (1995), Fiore and Bedwell (2011) described four types of team competencies for science teams and groups: (1) context-driven competencies specific to a given task and team; (2) team-contingent competencies that are relevant to a particular team but can be applied across various tasks; (3) task-contingent competencies that are relevant to a particular task, regardless of what team performs the task; and (4) transportable competencies, which can be applied across tasks and teams.
Cannon-Bowers et al. (1995) suggested that the first three types of competencies (specific to the task and/or the team) be developed through training for the team as a whole, while the more general “transportable” competencies be developed through education for individuals. Research on training and learning has shown that transfer of training is facilitated when the training context is similar to the context in which the trained skills will be applied (i.e., the workplace). Because the first three types of competencies are specific to a particular task and team context, Cannon-Bowers et al. (1995) suggested that training in these competencies be provided to intact teams (the specific team context) in their real work contexts or simulations of these contexts. Similarly, Kozlowski et al. (2000) proposed that if team members’ tasks are highly interdependent, training should focus on intact teams, while if their tasks are similar and can be simply pooled, team members can be trained as individuals.
Several recent meta-analyses attest to the effectiveness of team training in improving the knowledge, skills, and attitudes of teams (Salas et al., 1999; Salas, Cooke, and Rosen, 2008; Klein et al., 2009; Delise, Gorman, and Brooks, 2010). Salas, Cooke, and Rosen (2008) examined the impact of specific team training on various outcome measures (i.e., affective, cognitive, process, and performance) and found that team training had a moderate, positive impact on team process (ρ = .44) and performance (ρ = .39).
These findings were further supported by another team-training meta-analysis that found that, in general, team training had positive effects (Delise, Gorman, and Brooks, 2010). This meta-analysis suggests that training may be more effective for learning when individuals have the opportu-
nity to use the learned skills in the transfer environment. This is particularly promising for training of science teams and groups, suggesting that trainees could integrate the target skills into their daily activities to improve cognitive processes, such as deep knowledge integration, that leads to improved scientific performance (Salas and Lacerenza, 2013).
Team building is another intervention designed to improve overall team performance (Shuffler, DiazGranados, and Salas, 2011). Team building targets the interpersonal aspect of teamwork with particular emphasis on social interaction (Dyer, Dyer, and Dyer, 2007). Studies of team building have shown that it is not as effective as team training (Salas et al., 1999).
Fiore and Bedwell (2011) elaborated the work of Cannon-Bowers et al. (1995) to propose a competency framework to support research on professional development (training) of science teams (see Table 5-1).
In science teams, context-driven competencies are those related to a particular research project. Such competencies can be developed through training focused on project goals, research tasks, and methods. Teamcontingent competencies are those related to teamwork among these particular scientists and/or stakeholders and may be especially helpful to address
TABLE 5-1 Types of Team Competencies
|Relation to Task|
|Representative Science Team Competencies||Task-Specific||Task-Generic|
|Relation to Team||Team-Specific||
SOURCE: Adapted from Fiore and Bedwell (2011). Reprinted with permission.
challenges emerging from two features of team science—high diversity of team membership and high task interdependence. Team-contingent competencies can be developed through cross-training, in which individuals learn about the skills and duties of their teammates related to accomplishing scientific and/or translational tasks (see further discussion of cross-training below). For example, the Koch Institute for Integrative Cancer Research at Massachusetts Institute of Technology provides ongoing professional development opportunities to develop context-contingent knowledge of its particular research and translational mission and team-contingent competencies among its particular staff of life scientists, engineers, physicians, and other experts (see Box 5-1). Task-contingent competencies are those related to particular research tasks, such as experimental procedures. Finally, transportable competencies, useful across multiple science teams and/or larger groups, include such skills as mutual performance monitoring, giving and receiving feedback, leadership, management, coordination, communication, and decision making (Salas, Cooke, and Rosen, 2008).
This chapter now turns to a set of training strategies that show promise to address the coordination and communication challenges faced by science teams and larger groups. Many of these challenges can be addressed by developing team-contingent competencies, including “role knowledge”—understanding of the roles, tasks, skills, and knowledge each team member possesses. Coordination in science teams and groups can also be enhanced by developing context-driven competencies, including shared “mental models” (shared understandings of goals and tasks) among team members. Here, we discuss four research-based training strategies that show promise for enhancing coordination in science teams: cross-training, reflexivity training, knowledge development training, and team coordination training.
Cross-training can help members of science teams or groups develop both knowledge of the roles and capabilities of diverse team members and also shared goals. Cross-training was developed to teach “interpositional knowledge” within a team, defined as a form of shared knowledge that includes understanding of task and role responsibilities of all team members, as well as understanding of the factors that influence the team and shared expectations about how the team will respond to changing environmental situations (e.g., Cannon-Bowers et al., 1998; Cooke, Kiekel, and Helm, 2001; Hollenbeck, DeRue, and Guzzo, 2004). Teams without such knowledge often suffer from coordination and communication problems (Volpe et al., 1996). Cross-training has been shown to improve the development of team interaction and shared mental models, which led to improved coor-
Professional Development for Deep Knowledge Integration at the Koch Institute
The mission of the David H. Koch Institute for Integrative Cancer Research at Massachusetts Institute of Technology (MIT) can be briefly summarized as: “science + engineering = conquering cancer together” (see http://ki.mit.edu/ [April 2015]). This large group of scientists includes approximately 700 faculty, staff, and students within a 192,000-foot square building opened in the spring of 2011. Its research includes programs funded by the National Cancer Institute for multiinvestigator grants in the areas of systems biology and cancer as well as nanotechnology and cancer.
The institute’s core intramural faculty consists primarily of biologists and engineers who formerly worked in different MIT departments, along with a small number of physician-scientists who both treat patients and have laboratories at the institute, students, and postdoctoral fellows in all of these fields. Through its “Bridge” project, the institute links its investigators to many more physician-scientists at area medical centers. The confluence of these multiple disciplines leads at times to “messy, turbulent waters” and a tower of Babel situation, according to Institute Director Tyler Jacks. However, the institute members are beginning to better understand each other, partly through participation in multiple, structured professional development opportunities. As shown in Figure 5-1, they include
- The Friday Focus seminar series, where graduate students and postdoctoral fellows join faculty mentors in presenting research methods and findings to the entire institute staff. For example, one seminar was humorously titled “Attack of the Layer-by-Layer Nanoparticles: Co-delivery of Chemodrug and RNAi for Cancer Treatment.”
- Crossfire, a weekly educational series designed to bridge the biology/ engineering divide. The popular series was initiated by students and doctoral fellows, who both teach and attend the sessions in a peer-to-peer learning approach.
- A monthly lecture series, “The Doctor Is In,” which helps scientists and engineers understand cancer through talks by physicians.
- An engineering “Genius Bar,” created by postdoctoral fellows. Every 2 weeks, engineering fellows are available to answer questions on a specified topic.
- An annual retreat for all staff with hundreds of presentations by institute members along with poster sessions.
From the perspective of the literature on team training (Fiore and Bedwell, 2011; see Table 5-1), these seminars, lectures, and discussions aim to develop context-driven competencies related to the institute’s unique research and translational mission and team-contingent competencies, including knowledge of other institute members’ expertise and roles. The professional development opportunities provide forms of cross-training that may help biologists and engineers to better understand and appreciate each other’s skills, expertise, and duties related
FIGURE 5-1 Posters illustrate some of the Koch Institute’s professional development opportunities.
SOURCE: Presentation by Tyler Jacks to the committee, July 2013. Reprinted with permission.
to accomplishing shared research tasks and goals (see further discussion of cross-training below). As noted in Chapter 3, shared understanding of other team members’ expertise and roles, referred to as “transactive memory” has been shown to enhance team effectiveness.
SOURCE: Presentation to the committee by Tyler Jacks, director of the Koch Institute. See http://www.tvworldwide.com/events/nas/130701/default.cfm, click on “Why Team Science” [April 2015].
dination and backup behaviors, and, consequently, improved performance (Marks et al., 2002) and team decision making (McCann et al., 2000).
Three types of cross-training methods are commonly used: (1) positional clarification, in which individuals are told about the other positions on their team; (2) positional modeling, in which individuals are both told about the position and have the opportunity to observe or shadow the position, thus gaining a deeper understanding of the duties involved; and (3) positional rotation, in which individuals are given hands-on training in the other positions such that they are able to perform the role if needed (Salas, Cooke, and Rosen, 2008; Klein et al., 2009; Delise, Gorman, and Brooks, 2010). Positional rotation was shown to improve teamwork knowledge and overall team performance over more traditional procedural or rule-based training in a simulated team environment (Gorman, Cooke, and Amazeen, 2010).
Positional rotation of investigators is generally not practical within an interdisciplinary or transdisciplinary science team or larger group, as learning to perform another’s job would require obtaining an advanced degree in an unknown discipline. Nonetheless, more narrowly focused forms of cross-training, targeting the understanding of the roles, tasks, and expertise of team or group members, are feasible. Many of the courses and seminars offered at the Koch Institute are designed to help engineers and life scientists learn about others’ roles, tasks, and expertise through direct engagement with each other. They go beyond positional clarification, in which an outside trainer or facilitator tells team members about others’ roles, and are similar to positional modeling, in which the trainee observes or shadows a team member to learn about her or his role. For example, the engineering genius bar is an opportunity for life scientists, physicians, or other institute experts to directly observe engineers and ask questions about their work. Crosstraining supports the development of not only shared mental models (Marks et al., 2002)—a team process known to enhance team performance—but also “transactive memory,” or individuals’ knowledge of the specializations of team members. Research on new and hybrid cross-training approaches could help address the question of how much knowledge of other disciplines is sufficient for proficient engagement in team science.
Team Reflexivity Training
Team reflexivity training, if adapted and translated to science contexts, is likely to help science teams and groups develop positive processes such as team self-regulation and team self-efficacy, facilitating the complex coordination of work required for success. In a review of methods for improving science collaboration, Salazar et al. (2012) suggested that enhancing reflexivity in science teams can improve team creativity as well as integration of
individual member’s knowledge. As discussed in Chapter 3, the life cycle of a team has been conceptualized in terms of episodes of planning, action, and reflection. Team reflexivity training requires members to reflect on prior performance, considering which objectives were or were not met, the strategies used or the group processes engaged, and how performance could be improved in the future, with the goal of improving future interaction (Gurtner et al., 2007). Reflections are prompted by a series of questions for team discussion, without the use of a facilitator or trainer, making this form of training relatively brief and inexpensive. Gurtner et al. (2007) found that teams receiving reflexivity training developed shared mental models to a greater extent than a control, with a positive impact on collaborative performance. In another study, van Ginkel, Tindale, and van Knippenberg (2009) found that reflexivity training improved shared team understanding of tasks and decision quality.
Similar to reflexivity training, in self-correction training, participants are empowered to improve their performance by reflecting on past performance episodes and self-diagnosing areas for improvement. Whereas reflexivity training is generally applicable to any setting and can be facilitated by a series of questions without the use of a facilitator or trainer, self-correction training requires more initial training for proper use. Because self-correction training is more focused and specific than reflexivity training, it has the potential for greater benefits (Gurtner et al., 2007). Guided team self-correction, or team dimensional training, is a specific type of self-correction that was derived from an expert model of teamwork, and has been found to improve both taskwork and teamwork performance (Smith-Jentsch et al., 2008). As noted in Chapter 1, this approach has been generalized and found to improve team members’ shared mental models of teamwork across a variety of settings. It has been shown to increase performance and decrease errors in complex tasks such as naval submarine training simulations (Smith-Jentsch et al., 1998, 2008; Smith-Jentsch, Milanovich, and Merket, 2001).
Knowledge Development Training
Science teams and groups are composed of individuals with distinct sets of knowledge and expertise, which require integration to facilitate effective collaborative performance. This can be problematic given that research finds that different mental models of the task and the tendency to discuss commonly held information, as opposed to an individual’s unique information, reduce performance. To address these problems, Rentsch et al. (2010) conducted a study explicitly focused on team training for knowledge building. Teams of undergraduates were trained to engage in communicative processes that elicit the structure and organization of their knowledge
related to a team task designed by Navy Sea Air Land (SEAL) teams, as well as the assumptions, meaning, rationale, and interpretations associated with each member’s knowledge.2 The students used an external representation (i.e., an information board) that allowed team members to post, organize, and visually manipulate their knowledge related to the team task, more easily remember it, and draw attention to specific information as appropriate. The results showed that the knowledge-building training led to improved knowledge transfer (i.e., the exchange of knowledge from one team member to another), knowledge interoperability (i.e., shared knowledge that multiple team members are able to recall and use), cognitive congruence (i.e., an alignment or matching of team member cognitions), and higher overall team performance on the task (Rentsch et al., 2010).
In a follow-up study, Rentsch et al. (2014) tested a team-training strategy aimed at facilitating team knowledge-building in distributed teams. The authors found that teams trained to build knowledge, relative to untrained teams, shared more unique information, transferred more knowledge, developed higher cognitive congruence, and produced higher-quality solutions to a realistic problem-solving task.
Knowledge development training shows promise for improving collaborative problem-solving in science teams, by improving both knowledge building and knowledge sharing. However, other more general training strategies, such as reflexivity training and team development training, also improve knowledge building and knowledge sharing and, in addition, provide guidance in performance episodes.
Team Coordination Training
Team coordination training is a promising approach to facilitate the complex coordination of tasks required for success in science teams. This training was developed specifically to help teams modify responses based upon changes to their environmental situation. It focuses on helping teams adapt to the environmental demands of high-workload and time-stressed settings. This includes preplanning, information transmission, and anticipating information needs (Entin and Serfaty, 1999). It is primarily taught using vignettes to help teams recognize effective and ineffective teamwork. Practice and feedback are then provided in sessions where teams are able to apply what they have been taught and modify applications based upon errors. The goal is to turn explicit interaction factors that are thought to require effort on the part of the team (e.g., requests for information) into
2 Open communication about assumptions and meanings underlying one’s knowledge is also an element of the Toolbox intervention for interdisciplinary science teams and groups discussed later in this chapter.
implicit factors (e.g., providing information without being asked), in order to improve coordination. Although team coordination training was developed to help teams in contexts of high workload and stress, the competencies it develops (e.g., preplanning, anticipating information needs) are also suitable for teams in other contexts.
Gorman, Cooke, and Amazeen (2010) also explored a form of coordination training using methods described earlier in the cross-training section. The authors examined how to improve adaptability in teams through training that included disruptions to learned team coordination mechanisms. This involved, for instance, disrupting communication channels the team used to coordinate. Gorman, Cooke and Amazeen (2010) argued that this process-oriented training method helped teams deal with variability in coordination demands. Teams with disruption or “perturbation” training responded more adaptively to novel events than those with either cross-training or procedural training. The authors suggested that, similar to learning research on variability in practice, this helped teams generalize adaptive processes. By introducing coordination variability to the training, teams learned how to adapt their responses to changes in their environment and improve coordination during performance episodes. Science teams and larger groups face uncertainties that can arise from research findings (e.g., unanticipated results) or resource issues (e.g., loss of, or damage to, equipment; reduced grants) and hence might benefit from similar training approaches to increase their responsiveness to rapidly changing conditions.
Professional development designed specifically for science teams and groups is beginning to emerge, but only a few studies have examined its effectiveness for developing the targeted competencies or for improving performance. First, with support from the National Institutes of Health (NIH), the Northwestern University Center for Applied and Translational Sciences Institute developed an online training website, “TeamScience.net.” The website includes a series of learning modules, message boards, and linked resources that aim to enhance skills for participating in or leading interdisciplinary and transdisciplinary science teams or groups. Two expert users (an academic medical doctor and a medical librarian) reviewed the website, finding that it followed principles of instructional design for adult education, was easy to navigate, and used attractive audiovisuals to present lessons, along with links to additional information and outside websites (Aranoff and Bartkowiak, 2012). But research to date has not included careful study of the website’s learning goals and the outcomes for the users.
Second, the Toolbox Project (see http://toolbox-project.org [April 2015]), supported by the National Science Foundation (NSF), is a training intervention designed to facilitate cross-disciplinary communication in science teams and groups. O’Rourke and Crowley (2013) developed the Toolbox instrument to facilitate philosophical dialogue about science and the Toolbox workshop as a place for that dialogue. The instrument includes 34 probing statements accompanied by Likert scales to indicate the extent to which a respondent agrees with each statement. The statements are designed to elicit fundamental assumptions about science, including statements about ways of knowing (epistemologies), values, and the nature of the world. At the workshops, participants first complete the instrument and then engage in a facilitated dialogue lasting about 2 hours. At the end of the dialogue, they again complete the instrument. Data obtained from the workshop, including an audio file and pre- and post-dialogue reactions to the statements, are provided to the participants for analysis and reflection.
Although both the Toolbox instrument and the workshops are based on extensive theory and research and appear to target knowledge, skills, and attitudes supportive of interdisciplinary communication, to date there has been no empirical evaluation of whether participation in a Toolbox workshop leads to sustained improvement in cross-disciplinary dialogue after the workshop is over. In partial answer to this issue, Schnapp et al. (2012) analyzed data from a post-workshop survey that has been administered to 35 of the 90 teams and groups that have participated in a workshop. Just over half of those surveyed provided responses, and of these, 85 percent indicated that the workshop increased their awareness of the knowledge, opinions, or scientific approach of teammates, while 77 percent reported that the workshop had made a positive contribution to their professional development. A modified instrument for the health sciences was pilot-tested in two workshops with 15 participants, 10 of whom completed pre- and post-workshop questionnaires (Schnapp et al., 2012). Comparison of pre- and post-questionnaires revealed changes in about 30–40 percent of the items, related to motivations, research approaches, methods, confirmation, values, and reductionism, suggesting that the dialogue had met its goal of encouraging participants to thoughtfully consider other points of view.
Basic mastery of science concepts, methods, and perspectives provides the foundation for team science. In the 1960s and 1970s, when health sciences faculty experimented with interdisciplinary courses that focused on broad skills, curriculum committees and professional associations responded by mandating minimum levels of disciplinary knowledge and skills (Fiore, 2008). Reflecting such concerns, we first discuss science, technology,
engineering, and mathematics (STEM) education in this section of the chapter, before turning to a discussion of interdisciplinary education.
Historically, science education has rarely prepared future scientists with the knowledge and skills required for effective knowledge integration and collaboration within a science team or larger group. Elementary and secondary school science classes typically ask students to work alone, listening to lectures, reading texts, or taking tests designed to measure recall of facts. There are few opportunities to learn to collaborate effectively or understand science as a social and intellectual process of shared knowledge creation (National Research Council, 2006, 2007b). At the undergraduate level, students majoring in science and the related STEM disciplines take courses dominated by lectures and short laboratory activities that often leave them with major misconceptions about important disciplinary concepts and relationships (National Research Council, 2006, 2012b).
At the doctoral level, some students participate in science teams and groups, but continue to receive little or no guidance or instruction on how to be an effective collaborator. Students develop deep conceptual understanding of topics and methods within a discipline, and are trained in its unique perspectives, languages, and standards of evidence (epistemologies). As a result, they may consciously or unconsciously develop a negative perception of other disciplines (National Academy of Sciences, National Academy of Engineering, and Institute of Medicine, 2005). The hallmark of doctoral education is the student’s individual, unique, and original research, and teamwork at this stage may be actively discouraged (Nash, 2008; Stokols, 2014).
Collaborative Education in STEM Classrooms
New developments in K-12 and higher education potentially could enhance preparation for team science, developing both disciplinary and interdisciplinary knowledge and collaborative skills. The NRC Framework for K-12 Science Education (2012c) draws on decades of research showing that engaging students in science practices—such as asking questions, developing and using models, or engaging in argument from evidence—helps them master science concepts and facts (National Research Council, 2007b). Although students often work in small groups when engaging in these science practices, instruction has not been designed to integrate development of collaboration skills along with STEM concepts and skills.
Collaborative learning activities are also being tested in higher education. Research has shown that undergraduate learning of STEM is strength-
ened when students work collaboratively to solve problems, reflect on laboratory investigations, and discuss concepts and questions (National Research Council, 2012b). However, these approaches have not been widely adopted by college faculty, and, as at the K-12 level, they focus primarily on acquisition of STEM content and skills, with little attention to collaborative skills.
Gabelica and Fiore (2013a) reviewed studies of three group learning interventions in STEM higher education: problem-based learning, team-based learning, and studio learning. In all three approaches, faculty members present students with an authentic problem or assignment and students work in small groups to understand the issues at hand, gather relevant information, and develop solutions. All three approaches have been shown to enhance students’ understanding of targeted STEM concepts and skills under certain conditions (Gijbels et al., 2005; Strobel and van Barneveld, 2009), and a few studies of team-based learning also reported gains in students’ interpersonal and teamwork skills (e.g., Hunt et al., 2003). However, interpersonal and teamwork skills were seldom measured, partly because students were sometimes reluctant to rate their peers’ contributions to the team’s work (Thompson et al., 2007).
Gabelica and Fiore (2013b) recommended ways to address this gap, suggesting that developers of such interventions integrate insights from the organizational research on teams. This would involve, for example, assessing students’ development of interpersonal teamwork skills through self-ratings of interpersonal skills (Kantrowitz, 2005) and behaviorally oriented rating scales for self- and peer-evaluations of contributions to the team (Ohland et al., 2012).
Borrego et al. (2013) also recommended that developers of group learning interventions draw on the team’s research. In a two-phase study, the authors first reviewed 104 articles describing student team projects in engineering and computer science. They found that faculty assigned team projects to advance diverse learning goals, including teamwork, communication skills, lifelong learning, sustainability, and professional ethics. The student teams experienced team process challenges (e.g., conflict), and faculty members tried to address these challenges as they arose but were not aware of methods from the organizational literature that could be used to illuminate the very challenges they had sought to address. Second, Borrego et al. (2013) reviewed the organizational literature related to five team processes identified as important in the studies of student teams, clarified how these processes impacted student success, and developed theories of team effectiveness specific to engineering education.
Finally, research by Stevens and Campion (1994) has identified transportable individual competencies required for effective teamwork, showing promise for use within collaborative STEM education. These authors not
only explicated teamwork competencies but also developed and validated the Teamwork Test (Stevens and Campion, 1999) for measuring these competencies.
In sum, research to date has shown that carefully designed educational interventions that engage students in small group investigations, discussion, and problem-solving activities can support STEM learning, but has not yet examined the potential of such small groups to also serve as contexts for learning teamwork skills. Integration of concepts and methods from the organizational sciences with STEM education could redress this gap.
Interdisciplinary and Transdisciplinary Higher Education
Stokols (2014) observed that science teams and groups often address the coordination and communication challenges arising in interdisciplinary or transdisciplinary research by drawing on online resources and/or providing training on an ad hoc basis. He proposed that longer-term education is needed to prepare a generation of scholars capable of addressing complex scientific and societal challenges in collaborative, interdisciplinary or transdisciplinary research environments. Consideration of this proposal is informed by reflecting on the United States’ long history of interdisciplinary education, as well as more recent courses and programs focusing specifically on team science.
As the health sciences began to develop interdisciplinary programs in the 1960s (Lavin et al., 2001), researchers were prompted to address the communication and teamwork challenges inherent in these educational approaches (Hohle, McInnis, and Gates, 1969). This led to the creation of interdisciplinary internships and fellowships designed to help students learn to communicate across disciplines (Lupella, 1972) and highlighted the need for research and training related to the development of collaboration skills in team settings (Jacobson, 1974). Although interdisciplinary education grew over the following decades, knowledge of how to support development of collaboration and teamwork skills remained relatively static (Fiore, 2008).
Interdisciplinary education has grown rapidly over the past four decades (Lattuca et al., 2013). Between 1975 and 2000, the number of interdisciplinary majors at U.S. colleges and universities increased by 250 percent, a period when student enrollments increased only 18 percent. However, colleges and universities have been slow to support this shift toward interdisciplinary teaching and learning with supportive formal policies and practices. Klein (1996) called on universities to support faculty professional development in interdisciplinary teaching and to protect faculty from discipline-centric norms, such as tenure reviews that punish work outside one’s discipline. She suggested that such supports as mentoring, physical
space for collaborations, and cross-disciplinary training would help to develop new norms of interdisciplinarity. More recently, Klein (2010) argued that, to sustain interdisciplinary teaching and learning, institutional support must be consistent and embedded within the university culture.
Defining Competencies for Team Science
A critical issue is the lack of conceptual clarity about the learning goals of interdisciplinary and transdisciplinary education that aims to prepare students for team science. Researchers have proposed a variety of team science competencies as important learning goals for such education. We next discuss these competencies and provide a clustering of them in Table 5-2. More problematic is the lack of prospective experimental or quasi-experimental studies of learning outcomes, as the research has relied heavily on surveys, interviews, and archival analyses.
Building on an earlier framework by Stokols et al. (2003), Nash and colleagues (2003) delineated three types of core competencies for the transdisciplinary scientist: (1) attitudinal; (2) knowledge; and (3) skill-based. They proposed that all three types could be developed through graduate and postgraduate education, including coursework, seminars, and workshops taught by disciplinary and interdisciplinary faculty; mentoring by research supervisors in multiple disciplines; group work with other transdisciplinary trainees, such as a journal club; and a supportive institutional environment.
Using a consensus study of expert opinion, Holt (2013) identified a somewhat similar list of competencies for effective performance in team science contexts and recommended that they be developed in graduate education through interdisciplinary coursework and seminars along with team research and projects. Borrego and Newsander (2010) developed another list of competencies in a study of the NSF Integrative Graduate Education and Research Traineeship (IGERT) Program, which supports training of scientists for interdisciplinary team science. The authors grouped the diverse learning outcomes articulated across 130 successfully funded proposals, as follows:
- Disciplinary grounding: Although the awards are interdisciplinary by definition, a full 50 percent of proposals argued that graduate student trainees would gain grounding in a specific discipline.
- Teamwork: The most clearly articulated learning outcome, in 41 percent of the proposals, was that the proposed center would create a culture of teamwork.
TABLE 5-2 Competencies for Productive Participation in Team Science
|Values, Attitudes, and Belief-Based Competencies||Valuing Interdisciplinary or Transdisciplinary Collaboration||Attitudes that predispose one to integrate knowledge from a varied set of disciplines||Nash et al. (2003); Klein, DeRouin, and Salas (2006); Nash (2008); Fiore (2013); Stokols (2014); Vogel et al. (2014)|
|The beliefs that such efforts are necessary and can lead to effective outcomes|
|Societal and Global Perspectives||Belief that complex problems should be approached from a broad, multilevel perspective||Borrego and Newsander (2010); Stokols (2014)|
|Collaborative Orientation||Values that emphasize inclusion of multiple and diverse perspectives||Klein, DeRouin, and Salas (2006); Hall et al. (2008); Fiore (2013); Stokols (2014)|
|Understanding Other Disciplines||Understanding core theories and methods from other disciplines||Nash et al. (2003); Nash (2008)|
|Knowledge-Based Competencies||Disciplinary Awareness and Exchange||Awareness of assumptions of own discipline, engage colleagues from outside disciplines||Schnapp et al. (2012); Holt (2013); Lattuca et al. (2013); Stokols (2014)|
|Skills and knowledge to think across disciplines and synthesize varied concepts and theories|
|Processes of Integration, Integrative Capacity||Develop shared interdisciplinary vision, modify work based upon influence of others||Marks et al. (2002); Borrego and Newsander (2010); Salazar et al. (2012); Holt (2013)|
|Disciplinary Grounding||Cultivation of deep knowledge within one or more disciplines||Borrego and Newsander (2010)|
|Scientific Skills Across Disciplines||Use theories and methods of multiple disciplines||Gebbie et al. (2007); Vogel et al. (2012)|
|Methodology||Taking a methodologically pluralistic approach||Nash et al. (2003); Nash (2008)|
|Teamwork and Taskwork||Knowledge of resources and strategies to enhance teamwork as well as taskwork||McCann et al. (2000); Salas, Cooke, and Rosen (2008); Smith-Jentsch et al. (2008); van Ginkel, Tindale, and Van Knippenberg (2009); Borrego and Newsander (2010); Gorman et al. (2010); Holt (2013)|
|Interdisciplinary Research Management||Develop team skills to strengthen team structure and dynamics||Holt (2013)|
|Interpersonal/Skill-Based Competencies||Leadership||Build communication strengths, manage conflict, trust the value of teammates||Bennett and Gadlin (2012); Holt (2013); Ekmekci, Lotrecchiano, and Corcoran (2014)|
|Fruition||Presenting research at interdisciplinary conferences, partner with those in other disciplines on proposals||Holt (2013)|
|Interdisciplinary Communication||Active listening, oral and written, assertive communication Communicate regularly with scholars from other disciplines||Klein, DeRouin, and Salas (2006); Gebbie et al. (2007); Borrego and Newsander (2010); Fiore (2013)|
|Interact with Others Coordination||Engage colleagues from other disciplines Capacity to adapt flexibly and effectively to situational and intra-team challenges||Gebbie et al. (2007); Vogel et al. (2014) Entin and Serfaty (1999); Klein, DeRouin, and Salas (2006); Gorman et al. (2010); Fiore (2013)|
|Interdisciplinary Skills||Ability to consider and apply perspectives from outside one’s discipline||Lattuca et al. (2013)|
|Transdisciplinary Behaviors||The behaviors that support activities for integrating perspectives and working with others outside one’s discipline||Stokols (2014)|
|Intrapersonal-Based Competencies||Intellect and Self-Awareness||Broad intellectual curiosity, recognition of personal strengths and weaknesses with regard to interdisciplinary research||Hall et al. (2008); Holt (2013)|
|Reflective Behavior||Ability to recognize when one’s general approach, or a specific problem-solving approach, needs to be changed||Lattuca et al. (2013); Stokols (2014)|
|Critical Thinking||Critical awareness about one’s own potential disciplinary biases in collaborative situations||Borrego and Newsander (2010); Hall et al. (2012a); Vogel et al. (2014)|
- Integration: Thirty percent of the proposals argued that their graduate programs would encourage students to integrate concepts from relevant disciplines.
- Societal and global perspectives: Twenty-four percent of the proposals noted that they would encourage students to consider societal and global issues.
- Interdisciplinary communication: Twenty-four percent of the proposals noted that their projects would emphasize the importance of interdisciplinary communication.
Borrego and Newsander (2010) also found that scientists, engineers, and scholars in the humanities had different views of “integration.” For scientists and engineers, “teamwork” was fundamental, whereas scholars in the humanities considered “critical thinking” as more central. The authors suggested that because critical reflection on disciplinary inconsistencies and limitations is a particular strength when solving complex interdisciplinary
problems, scientists and engineers should incorporate critical thinking as a goal of interdisciplinary education.
Engineering students are often assigned to work in interdisciplinary teams, and Lattuca, Knight, and Bergom (2013) developed a self-report measure of interdisciplinary engineering competence, including three scales: interdisciplinary skills, reflective behavior, and recognizing disciplinary perspectives. Importantly, the scales do not include any measures of teamwork or interpersonal skills. Lattuca, Knight, and Bergom (2013) caution that the scales are preliminary and that they were unable to evaluate their construct validity (their relationship to the target competencies), “because direct measures of interdisciplinary knowledge and skills do not exist” (p. 737).
Gebbie et al. (2007) identified competencies for transdisciplinary health research. Using a Delphi technique to elicit information from several groups of experts in interdisciplinary research and education, they arrived at 17 statements describing what a well-trained scholar should be able to do when participating in interdisciplinary research. The statements were grouped into three categories: conduct research, communicate, and interact with others.
As discussed above, Cannon-Bowers et al. (1995) suggested “transportable team competencies” as a focus for educational programs to develop the kinds of competencies that can be applied across different tasks and teams. Building on this, as well as a framework of interpersonal skills created by Klein, DeRouin, and Salas (2006), Fiore (2013) developed a framework of transportable interpersonal competencies for team science. This framework specified the forms of active listening, oral and written communication, assertive communication, relationship management competencies, coordination, interdisciplinary appreciation, and collaborative orientation that support effective collaboration in science. Fiore suggested that these competencies be integrated as learning goals for interdisciplinary education to support team science.
Stokols (2014) conceptualized a broad intellectual orientation for transdisciplinary team science including values, attitudes, beliefs, skills and knowledge, and behaviors (see Table 5-2). Both Stokols (2014) and Misra et al. (2011a) emphasized the role of mentors in graduate education, noting that mentors who encourage the acquisition and synthesis of a broad knowledge base can help students acquire the skills and attitudes foundational to transdisciplinary work. Stokols (2014) also suggested that, when students are trained in institutions that engage them in authentic team science research activities focused on real-world problems, “they are better able to avoid the conceptual biases associated with disciplinary chauvinism and the ethnocentrism of traditional academic departments” (p. 66).
In sum, many authors have proposed various competencies for team science and educational strategies to develop these competencies, and there
are areas of overlap and agreement within this variety. However, the research to date has not identified a common set of agreed-on competencies that could serve as targets for design of educational or professional development courses.
Research on Educational Interventions for Team Science
There have been only a few empirical analyses of educational strategies aimed at preparing individuals for team science. These educational strategies vary, including programs implemented within individual schools or universities as well as larger, federally funded education programs. In addition, the research to date has not examined how acquisition of the targeted competencies may enhance the effectiveness of science teams.
Graduate Education for Team Science
The University of California, Irvine’s School of Social Ecology offers a doctoral seminar specifically developed to expose students to a broad range of relevant disciplines. To examine the influence of the seminar, Mitrany and Stokols (2005) conducted a content analysis of doctoral dissertations produced by the school, reporting that the dissertations provided evidence of an interdisciplinary orientation reflected, for example, in the multidisciplinary composition of their faculty committees and the cross-disciplinary scope of their research topics, conceptual frameworks, and multimethod analyses.
Carney and Neishi (2010) conducted an evaluation of the IGERT Program described above, using surveys and data from IGERT graduates and a control group of doctoral graduates from similar academic departments that did not participate in the program. In comparison to the non-IGERT graduates, a higher percentage of IGERT graduates reported that they drew on at least two disciplines in their dissertation research and obtained their doctoral degrees in less time (thanks to the program’s financial support). Contrary to some previous authors who warn that interdisciplinary doctoral students may face challenges in the discipline-based academic job market (e.g., Nash, 2008), the IGERT graduates reported that their interdisciplinary research training and the program’s professional networking opportunities gave them a competitive edge in the job market. They reported less difficulty acquiring their first jobs than the non-IGERT graduates. At these jobs, they were significantly more likely than their non-IGERT peers to conduct research or teach courses that require integration of two or more disciplines.
In a separate study of the IGERT Program, Borrego and colleagues (2014) sought to identify longer-term outcomes of the traineeships for the
host universities as well as the trainees, by interviewing faculty and administrators at a small number of institutions. The interviewees reported overcoming barriers to successful implementation of the interdisciplinary doctoral training program through, for example, changes to eligibility criteria for advisers so that faculty from varied departments could serve as a doctoral student’s adviser. In addition, departments changed their policies to reward faculty for advising outside their department, and some institutions expanded eligibility for fellowships so that students from interdisciplinary programs could compete for the awards. In addition, many programs created interdisciplinary courses or seminars and required that students participate in team research and take laboratory classes from different disciplines.
The National Cancer Institute’s Transdisciplinary Research on Energetics and Cancer I (TREC I) project sought to develop three types of competencies for graduate students (Vogel et al., 2012):
- scientific skills, including educational grounding in two or more disciplinary perspectives and skills for integrating and synthesizing approaches across disciplines;
- intrapersonal skills, including positive attitudes, values, and beliefs about the transdisciplinary approach and critical awareness of the relative strengths and limitations of all disciplines (referred to as a transdisciplinary orientation); and
- interpersonal skills for collaborating and communicating across disciplines, such as the ability to use analogies, metaphors, and lay language in lieu of discipline-specific jargon and willingness to engage in continual learning.
The four TREC centers implemented a variety of training activities to develop these competencies, including interdisciplinary research courses, journal clubs, and writing retreats to develop skills in collaborative writing and research. Many centers also provided co-mentoring and multi-mentoring to expose trainees to multiple disciplinary perspectives, and a cross-center working group developed additional training activities.
Multiple mentors were expected to play a key role in developing the three types of competencies, by teaching trainees about the concepts, theories, and methods of the different disciplines; facilitating learning of interpersonal skills for transdisciplinary research; and providing support for career advancement (e.g., the mentors would provide visibility to and coach the trainee). The “multi-mentoring” approach was also expected to provide social support and role modeling. However, each TREC center was allowed to develop its own training program, and the study found that only about 60 percent of trainees had two or more mentors.
An analysis of these training activities found gains in all three types of competencies, including students’ attitudes toward working across disciplines, ability to work across disciplines, and scientific competency. Importantly, the trainees also improved in scholarly productivity, as measured by number of publications/presentations and number of collaborative authors. Multimentoring experiences were associated with greater transdisciplinary orientation and positive perception of one’s center (Vogel et al., 2012).
Undergraduate Education for Team Science
Few studies have examined the goals and outcomes of interdisciplinary undergraduate programs focusing on team science. One example was a study of the University of California, Irvine’s Interdisciplinary Summer Undergraduate Research Experience Program, which aims to develop students’ ability to integrate research concepts and methods. Misra et al. (2009) examined curriculum strategies (such as the use of team projects, research, or journal club meetings), interdisciplinary processes (such as student participation in team projects), and student products (completed projects, papers, and course grades) for a group of participants. Over the course of the program, participants developed more positive attitudes toward interdisciplinary research and participated in interdisciplinary research activities more frequently. In comparison with another group of students who participated in a different research fellowship program that did not include an interdisciplinary component, the participants showed no significant difference in student products, but a higher level of engagement in interdisciplinary collaborative research. Further, team-focused projects were found to be instrumental to these changes.
In light of Borrego and Newsander’s (2010) suggestion that critical thinking is valuable for interdisciplinary collaboration in science and engineering, a recent study by Lattuca et al. (2013) focused on this competency. In a longitudinal study of about 200 students, the authors compared students majoring in traditional disciplinary programs with those participating in interdisciplinary programs, using existing assessments of critical thinking, need for cognition, and attitudes towards learning. They found no significant differences in levels of these competencies between the two groups that could be attributed to major or structure of the program.
The Role of Mentoring for Team Science
The research discussed above consistently identifies mentoring as a crucial component of interdisciplinary education for team science, but only a few programs focus specifically on mentoring. For example, NIH’s Building Interdisciplinary Research Careers in Women’s Health Program is designed
for junior faculty interested in advancing research in women’s health. The program establishes mentoring teams to provide the young faculty members with multiple perspectives on a range of scientific and career issues. A recent study showed that a majority of scholars in the program had applied for competitive grants after completing the training and that approximately half were successful (Nagel et al., 2013).
In 2010, NSF adopted a new policy requiring that requests for funding of postdoctoral researchers include a postdoctoral researcher mentoring plan. Implemented in part to advance NSF’s two core strategies of fostering the integration of education and research and expanding the participation of groups and institutions that have been underrepresented in science, the plans must describe mentoring activities, such as career counseling, training in preparation of grant proposals and publications, and “guidance on how to effectively collaborate with researchers from diverse backgrounds and disciplinary areas” (National Science Foundation, 2014b). Recent reports, although anecdotal, suggest that reviewers of NSF proposals may be placing increased weight on this requirement (Flaherty, 2014).
Currently, however, mentoring, and especially interdisciplinary mentor-ship, is too often lacking for students and scholars. In a recent survey on the “Global State of Young Scientists,” the unavailability of mentoring was one of the top four career obstacles identified (Friesenhahn and Beaudry, 2014). Survey responses indicated that junior scientists are not explicitly taught how to train and supervise students and postdoctoral fellows, but are expected to learn by experience.
In this section, we consider how the research reviewed in this chapter may help guide professional development, training, or education for team science as a way to address the communication and coordination challenges that emerge from the key features that create challenges for team science.
High Diversity of Membership
The challenges of communication and interpersonal interactions raised by high diversity of team membership can be addressed in part with crosstraining and other types of training focusing on team-specific competencies, to help team members better understand and appreciate the varied knowledge and roles of different team members. These challenges also can be addressed through interdisciplinary educational seminars that expose team members to scholars from different disciplines, such as those offered by the Koch Institute or through structured approaches such as the Toolbox work-
shops described above. In addition, professional development or education for team science could focus directly on development of interpersonal skills such as “active listening” with the goal of ensuring that inputs from those in different disciplines are understood.
Deep Knowledge Integration
As noted in Chapter 1, science teams and groups that seek to deeply integrate, or even transcend, the knowledge of individuals who may have different goals, assumptions, and languages often encounter communication and coordination challenges. Professional development focused on developing shared understanding of each member’s knowledge—such as cross-training, knowledge sharing training, and coordination training—may help to address these challenges. Education or professional development devised to illustrate larger connections across disciplines (both conceptual and methodological) also would help foster knowledge integration.
Although training to develop shared knowledge of fellow team or group members’ knowledge and skills can help to overcome the communication and coordination challenges raised by large size, this training may have to be relatively shallow. For example, cross-training may focus on positional clarification (knowledge of other members’ roles), rather than deeper understanding of other members’ knowledge, skills, and tasks, both because of the large number of members and because it is not practically possible to quickly develop deep understanding of an unfamiliar discipline. As a first step, leaders of large groups may consider engaging training experts to identify the amount of “interpositional knowledge” necessary to support behavioral coordination across the team.
Goal Misalignment with Other Teams
Lack of goal alignment with other teams may result partly from team members’ lack of awareness of common goals and partly from organizational factors that are beyond the scope of team training. Training or professional development can be designed to increase awareness of the common goal and how the goals of the varied subteams are linked to that goal. In addition, this challenge can be addressed through reflexivity training. Teams that reflect on prior performance episodes can develop knowledge of when goal alignment and/or misalignment with other teams is affecting their interactions and performance. Educational interventions that include group activities, such as problem-based learning and team-based learning,
also could introduce the concept of goal alignment to help students learn how to manage goal conflicts that often arise between different science teams.
Permeable Team and Group Boundaries
Permeable boundaries create a need for the context-driven, team-contingent, and task-contingent competencies shown in Table 5-1. In terms of the context, team or group members who are new to a project would need training in the project’s scientific and/or translational goals. From the task standpoint, new members may require training in particular research methods or analyses to accomplish research tasks. From the team standpoint, transitional membership creates a gap in team-specific knowledge, as a new member may not understand teammates’ expertise and roles. Such a gap could be addressed by cross-training or knowledge development training.
Geographic dispersion of team members necessitates training to develop team or group members’ understanding of each other’s expertise, roles, and context-driven and team-contingent competencies. Cross-training or knowledge development training may help to provide this understanding, thus facilitating coordination. However, because dispersion hinders acquisition of this understanding, training focused on development of team cohesion or team self-efficacy might also be beneficial. Reflexivity training can also be used to identify when and where proximity is creating problems for the team.
High Task Interdependence
The high level of interdependency within science teams and groups creates a need for both context-specific and team-specific knowledge. Because one or more members is likely to have the expertise needed to accomplish each piece of the research project (e.g., expertise in statistics), knowledge of different team or group members’ expertise can facilitate coordination, supporting team effectiveness (Kozlowski and Ilgen, 2006). To develop context-specific competencies, training should focus on task-specific knowledge and skills. To develop team-specific knowledge, reflexivity training is a promising method. Both training strategies can support the deep integration of team members’ knowledge needed to achieve the scientific and/or translational goals of the project.
Research on teams in a variety of contexts outside of science has been applied to develop training strategies, shown to improve team processes and effectiveness. Several research-based training strategies show promise to enhance communication, coordination, and knowledge integration in science teams, overcoming the challenges that emerge from diverse membership, large sizes, high task interdependence, and other features of team science. The committee expects that translation and application of these strategies to create professional development programs for science teams would enhance the effectiveness of these teams. Professional development programs for team science are beginning to emerge, and these programs would benefit from translation and application of the strategies shown to enhance effectiveness in non-science contexts.
CONCLUSION. Research in contexts outside of science has demonstrated that several types of team professional development interventions (e.g., knowledge development training to increase sharing of individual knowledge and improve problem solving) improve team processes and outcomes.
RECOMMENDATION 2: Team-training researchers, universities, and science team leaders should partner to translate, extend, and evaluate the promising training strategies, shown to improve the effectiveness of teams in other contexts, to create professional development opportunities for science teams.
The TeamSTEPPS Program illustrates the approach the committee recommends to improve the training and performance of science teams. TeamSTEPPS extends and translates research findings on team effectiveness in aviation to create health care team-training practices with the goal of improving health care performance. The program was developed in response to the Institute of Medicine (1999) report To Err Is Human: Building a Safer Healthcare System, which identified the need to improve team performance in medical settings as one of several steps recommended to reduce medical errors and improve health care. As described by Alonso and colleagues (2006), the program’s developers reviewed more than 20 years of research on teams and team performance to identify critical competencies needed for effective teamwork and translate them for health care contexts. The list of competencies was then converted into a framework of trainable team skills, and training strategies were developed to strengthen these skills.
Although research has demonstrated that training for current team members can increase team effectiveness, educational programs designed
to prepare students for future team science have only recently emerged and have not yet been systematically evaluated. Further work is needed to more clearly specify the competencies needed for team science and to develop assessments of these competencies; such research would clarify learning goals, an important step toward enhancing learning outcomes. To date, there has been little empirical evaluation of which educational activities are most effective for developing particular competencies, nor whether, and to what extent, acquisition of these competencies contributes to the effectiveness of science teams or larger groups.
CONCLUSION. Colleges and universities are developing cross-disciplinary programs designed to prepare students for team science, but little empirical research is available on the extent to which participants in such programs develop the competencies they target. Research to date has not shown whether the acquisition of the targeted competencies contributes to team science effectiveness.