What Makes a Broker a Broker?

The authors are Co-directors of the Michigan School Program Information (MiSPI) Project

Brokers play an important role facilitating the transfer of research evidence between those who produce it (i.e. researchers) and those who put it into practice (i.e. practitioners), and thus in closing the research-practice gap. The role of brokers is now widely acknowledged in the growing literature on the use of research evidence, but what makes a broker a broker? There appear to be two broad approaches to answering this question—one that focuses on structure, and another that focuses on function—but definitions are often left implicit. Really understanding how brokers work, however, requires being explicit about what they are and reconciling their different approaches.

The structure of brokerage

The structural conception of brokerage contends that what makes a broker a broker is where they are located within a network of communication and influence. In a caricatured sense, researchers produce research, practitioners use research, and everyone else is or could be a broker that links these two communities. That is, brokers are the people and organizations that neither produce nor use research, but instead are positioned in a network of communication and influence to make it possible for practitioners to acquire and use research, and for researchers to learn about practitioners’ experiences. Within this structural conception of brokerage there are several approaches.

Geodesic brokers are the people and organizations that link researchers (R) and practitioners (P) most efficiently. But, information rarely follows the most efficient route, so this definition can be too restrictive. A more inclusive approach seeks to also identify non-geodesic brokers that link researchers and practitioners, even if these links are indirect or inefficient. Geodesic and non-geodesic brokers can be located by tracing the paths that pieces of information follow using a small world design. We have proposed that this design offers a way to understand the nature and severity of the research-practice gap, as well as insights into ways it can be closed.

The triadic approach takes a narrower view by focusing on just three key actors at a time: a source of information (S), a broker (B), and a recipient of the information (R). This approach aims to identify different types of brokers based on the settings where these three actors are found. For example, brokers who come from the same setting as recipients can act as gatekeepers that block sources’ access to recipients. Alternatively, brokers who come from the same setting as sources can act as representatives to recipients on behalf of sources. Finally, brokers who come from different settings than either sources or recipients can act as liaisons between these settings. In our work with public school administrators, we have observed these three types involved in the transfer of research evidence.

The function of brokers

The functional conception of brokers contends that what makes a broker a broker is what they do. Brokers can serve multiple and different functions in communicating information to practitioners and policymakers. Moreover, these functions can be roughly ordered by how much brokers insert themselves into the process, in contrast to the structural conception that focuses on where brokers insert themselves.

Brokers may selectively pass along certain pieces of information they deem credible or useful, and thus perform a filtering function. For example, a broker might share an article on the effectiveness of grade retention with a school principal, but chose not to share an article on new practices for teacher evaluation. Filtering information involves some insertion of brokers into the process because they must make judgments about which information to share.

Because the type of information produced by researchers is often not formatted for practitioners, brokers can be helpful by synthesizing this information. In some cases, synthesizing may involve aggregating multiple pieces of information in a convenient place, for example, in clearinghouses like What Works Clearinghouse and the National Registry of Evidence-Based Programs and Practices that serve as repositories of research for educators. In other cases, it may involve translating raw research findings as reported in journal articles into more readily digestible forms like executive summaries and program profiles. Synthesizing requires more insertion of brokers because they must review, interpret, and alter the content and format of the information presented.

Finally, brokers sometimes get involved by advocating for practitioners’ use of particular pieces of information. For example, recent Foundation-funded research on research use in charter school reforms has found that advocacy organizations often use information to push their own agendas to policymakers. Advocacy involves a high degree of insertion on the part of brokers as they actively use information to advance their own goals.

Linking structure and function

Within these two conceptions, there is significant fuzziness. For example, from the structural perspective, the same person might act as a gatekeeper one day, but as a representative the next. Similarly, a single organization might be simultaneously engaged in filtering and advocating when it picks and chooses which pieces of information best advance its agenda.

Although researchers studying brokers are already aware of this kind of fuzziness, they still often adopt either a structural or a functional conception of brokers. But, are these two ways of thinking about brokers really so different? Probably not. In many cases, the structure of a network and the function of its participants are tightly coupled. For example, in a hierarchical workplace, the person at the top is well-positioned to give orders, while the people at the bottom are well-positioned to follow orders. But if things were reversed—if the people at the bottom all give orders and the person at the top tried to follow them—chaos would ensue. Following this logic of a tight coupling between structure and function, certain structures of brokerage likely facilitate certain functions of brokers.

The function of filtering the information that reaches a recipient may be easiest for a broker occupying the position of a gatekeeper. Because a gatekeeper broker comes from the same setting as the information’s recipient, they already “speak the same language,” allowing the broker to receive information from outside sources and make informed judgments about what is important or worthwhile to pass along.

The function of synthesizing information may be easiest for a broker occupying the position of a liaison. Researchers have highly specialized skills in conducting research and generating evidence, while practitioners likewise have specialized skills in delivering services. Because liaison brokers come from a different setting, they have the freedom to develop a different set of specialized skills: synthesizing complex research findings into actionable and digestible forms, or translating practitioners’ experiences into data.

Finally, the function of advocating for the use of a particular piece of evidence or evidence-based practice may be easiest for a broker occupying the position of a representative. Because a representative broker comes from the same setting as the information’s source, they share a goals and agenda, but unlike the source itself, a representative broker also has direct contact with the information’s intended recipient and user.

This sketch of how the structures of brokerage can be linked to the functions of brokers highlights that these two different conceptions of brokers are intimately related. Although there is probably not such a simple one-to-one matching of structure and function, the notion of a tight coupling between the two can bring greater clarity to the question: What makes a broker a broker? By conceptualizing brokerage structures and broker functions as related, it is possible to integrate two existing bodies of literature on brokers and knowledge transfer. This approach can also help us find ways to leverage brokers more effectively to improve the use of research evidence. For example, in situations where the research evidence is particularly complex, and thus where synthesis is among the most important functions for brokers, efforts can focus on supporting individuals in liaison-type positions. In contrast, when the key challenge to use is the sheer volume of research evidence and other information, and thus where filtering is a critical function of brokers, supports can instead focus on those in gatekeeper-type positions. But across all these types of situations, recognizing that structure and function are complementary, not competing, ways of understanding brokerage will be important to facilitating the use of research evidence.