Establishing an AHRQ Learning Collaborative, A White Paper

Theoretical Frameworks

 Need for Collaboration in Health Services Research

According to Gray,1 collaborations are formed when two or more stakeholders invest their resources (e.g., talent, information, money), to solve problems that they could not solve as individuals. Collaborations are defined as:

...a process in which autonomous actors interact through formal and informal negotiation, jointly creating rules and structures governing their relationships and ways to act or decide on the issues that brought them together; it is a process involving shared norms, and mutually beneficial interactions... 2

Put another way, a collaboration is "a mechanism for collective action."3 For researchers, collaboration provides an opportunity to share individual intellectual strengths and expertise and to collectively expand the boundaries of knowledge through joint production. Irrespective of the intended outcomes of the collaboration, the "individual-to-collective" process is critical to success both theoretically, because individuals "are exposed to diverse kinds of relevant and legitimate information"4 and practically because these individuals are likely competing with one another for resources (i.e., funding). Thus, collaboration is a mechanism for increasing understanding of and appreciation for others' perspectives (through the joint production process) and doing so in a more efficient manner. Collaborations have become necessary for organizations performing complex work, with emerging technologies and rapidly changing environments.2

Central to this concept of collaboration is knowledge translation and knowledge transfer. Knowledge translation involves the exchange, assessment, review, and utilization of evidence-based research in organizations.5 Distributing knowledge by traditional mechanisms, such as publications and peer-reviewed journals, however, creates a gap between "what we know" and "what is being done."6 Sometimes, good scientific papers "get lost in the flood of bad papers."7 Knowledge transfers across disciplines, organizational boundaries, or domains tend to be "sticky" and difficult, as opposed to fluid. Knowledge "stickiness" can occur because the source of knowledge is unreliable, the source is not motivated to support transfer, or the recipient is unable to identify, value, and apply new knowledge.8 Knowledge translation is particularly difficult, costly, and time consuming in the field of medicine and HSR.

However, HSR is fast changing, ever expanding, multidisciplinary, and multi-institutional. HSR is housed in and supported through research universities (especially Schools of Medicine and Public Health), academic medical centers, research institutes, and government agencies (like AHRQ and the National Institutes of Health [NIH]), as well as a multiplicity of professional associations such as AcademyHealth. If we look at the landscape (or knowledge map) of Federal training programs in comparative effectiveness research developed by Tim Carey, we find we find a case in point (Appendix B). There are 55 (Clinical and Translational Science Awards (CTSAs) and 14 Evidence-based Practice Centers, as well as DeCIDE II, K12, and NIH KM1 programs. NIH funds HSR training programs and, in addition to supporting 28 T32 programs offering both pre- and postdoctoral training, AHRQ also supports (with ARRA [American Recovery and Rehabilitation Act of 2009] funding), eight K12 and seven T32 programs specific to comparative effectiveness research. All of these programs play important roles in seeking and creating knowledge that advances the technical, social, and political missions of HSR. But how can we become more adept at sharing and transferring knowledge when the organization of knowledge is separated by programs, disciplines, and otherwise-gated communities?

Although the theoretical benefits of collaboration should exceed the sacrifices, studies report that value is created only if a collaboration can be sustained over time.9,10 Successful and sustainable collaborations are not self-managing; rather, they require well-explicated processes and structure and capacity building.2 Capacity solves two problems: one, how to foster mutually beneficial relationships and effective participation without overloading the time and attention of members, and two, how to set common goals and work together to carry out activities related to those goals.10 

A useful framework for thinking about the social requirements of a collaboration come from Fair Process Theory.11,12 This framework (Figure 1) conceives collaboration as a periodic cycle of interactions. The model depends on the relational and rational processes over time. When people are deeply engaged in framing and defining a situation, exploring alternatives, and understanding the rationale for all decisions, they will commit to a course of action as long as they are able to negotiate "minimal and congruent" expectations.9 Evaluating commitments is based on their perceptions about fairness and reciprocity. If commitments are not reciprocal, people will renegotiate or reduce their participation.

In the next section, we review some of the literature relevant to organizing and sustaining collaborations.

Return to Contents 

Communities of Practice: A Mechanism for Organizing and Sustaining Collaboration

One novel perspective on knowledge translation and collaboration comes from the work of Jean Lave and Etienne Wenger.13 According to Wenger and Snyder, 14 today knowledge only exists in learning networks or what are called "communities of practice."

Communities of practice are groups of people who share a concern or a passion for something they do, and learn how to do it better as they interact regularly... 15

Wenger argues that communities of practice have three characteristics: a domain, a community, and a practice. First, there is a shared identity, defined by a domain or area of interest. Members are committed to that domain. Second, a community is defined by members who help and share information. When help is given, it "enables another person to solve a problem, to accomplish something, or make something easier."16 Help is a social process that runs to the heart of any successful collaboration.2,17 Third, practice involves practitioners who interact regularly and who develop and share tools, concepts, resources, and capabilities for dealing with recurring problems and opportunities. To enable "reflective practitioners," some face-to-face interactions are essential.

Community of Practice theory argues that many different experts in widely dispersed universities are doing their own research, using their own methods, and making scientific progress much more slowly than if everyone worked together. Organizations that can source external knowledge via networks and collaborative partners build innovative products and effective programs faster than internally focused organizations.18,19

According to Cross and Parker,20 collaborations facilitate two major processes: one, making sense out of the opportunities and problems in the environment, and two, coordinating responsive actions. Although the type of collaboration depends on the situation, the nature of the work and the objectives require building a network of relationships. Two network archetypes have been noted:

  1. Customized response networks that can define problems rapidly and perform sense-making activities.
  2. Routine response networks that can deal with well-defined problems and solutions.

When problems are very complex, ill-defined, require sourcing knowledge from multiple disciplines or locations, and require different levels of expertise, groups can outperform individuals. Academic fields and industries driven by innovation and science have developed a capacity to source knowledge by learning from collaborations and learning how to collaborate with external partners. 

Recently, an unprecedented number of organizations have blossomed into communities of practice to accelerate scientific and other advances. One successful model for developing collaboration was developed by MindTree, an information technology company.21 MindTree wants to deepen collaboration and learning by creating communities of practice. They conceptualize knowledge management and communities of practice as a developmental pyramid that starts with common interests and can ultimately build capacity. In Figure 2, we show how this might work to build a collaborative among the AHRQ programs. The model in Figure 2 works as follows. When AHRQ trainees or faculty from two or more programs share interest in an issue, topic, data set, method, or theory, a potential community of interest emerges. At this basic stage, there are no objectives and not much information sharing or impact. A social network is activated, and stage two develops when there is shared learning across programs via interpersonal communication, activity, and interactions. At this stage, trainees and/or faculty have matured beyond shared interests and have discovered a way to build group competency across programs. The third stage develops a capability among several programs by transferring data, methods, tools, and best practices. In the final stage, not only has a national training capacity been established in a specific area, there also are bridges to knowledge and experts in the world who have become part of this collaboration.

In conclusion, the literature notes that one key to knowledge translation and transfer is engaging researchers and users in collaborative efforts. In theory, teams can outperform the "smartest" individuals or groups when multiplex tasks require a wide range of skills, experiences, and ideas.22 The opposite can also be true. Tom Malone, who has pioneered research on collective intelligence at MIT, has bemoaned, "Having a bunch of smart people in a group doesn't necessarily make the group smart."23 Group dynamics can lead to inferior performances well below the poorest performing members of a group. The Center for Collective Intelligence at MIT frames the challenge of collaborating as, "How can people and computers be connected so that collectively they act more intelligently than any individuals, groups, or computers have ever done before?"24 The literature suggests that the relational aspects of collaborating are critical. Thomson and Perry2 find that there must be complementarity and mutual benefits. Since trust is a key ingredient, collaborations cannot be forced. It takes time for the filaments to connect.

Page last reviewed October 2014
Page originally created March 2012
Internet Citation: Theoretical Frameworks. Content last reviewed October 2014. Agency for Healthcare Research and Quality, Rockville, MD.