Table summarizes the critical steps involved in the Brokered Dialogue method.
Critical Steps inBrokered Dialogue
The method begins with the selection of a socially significant issue or controversy in which there is a lack of dialogue among the relevant stakeholders. Participants are selected based on the particular perspective they are anticipated to bring to the dialogue (purposive sampling) [25
]. For example, we might conduct a Brokered Dialogue
about seasonal influenza vaccination among healthcare workers, since vaccination rates remain low worldwide despite explicit recommendations that they be vaccinated [26
]. We might begin by interviewing frontline practitioners who had and had not been vaccinated and perhaps an occupational health clinician involved in promoting staff vaccination and surveillance at a given institution.
Once participants have been identified, in-depth open-ended filmed interviews are conducted with each participant in order to elicit their experiences and perspectives on the topic. Drawing on interactive narrative methods, the positionality or ‘footing’ of given participants is recognized as important [27
]. Based on our preliminary thematic analysis, and a combination of theoretical and sequential referral sampling [25
] we might decide to interview representatives of public health agencies engaged in establishing influenza vaccination policies for healthcare organizations, representatives from the vaccine manufacturer, and hospital administrators.
Data collection and data analysis occur in conjunction with one another, in a recursive process [17
]. Analysis of introductory interviews combines the review of filmed footage and written transcripts to identify codes, themes and concepts, as well as patterns of ideas that may serve as points of intersection for developing dialogue [15
]. Comparisons within and across interviews are made (‘constant comparison’) [32
]. A non-finalizing dialogic approach to analysis is adopted from this initial stage, whereby selections of the first interviews are compiled and compared, and points of divergence and convergence on the research topic identified [35
]. The potential for a multiplicity of interpretations is recognized [35
]. Based on this initial analysis, first ‘clips’a
of the individual interview films are assembled, highlighting each participant’s perspective, and how it may fit with others [2
]. Interviewees participate with us in editing their own film clips to ensure:
(a) that they accurately reflect their perspectives; (b) that they reflect a balance between the general narrative of the participant’s perspective, and specific points of contention; and (c) that they feel comfortable showing them to others. This iterative editing process is the visual equivalent to the textual ‘report’ generation described by Frank in textual dialogic analysis, creating “an opening … to further representations” [35, p. 44].
Following the creation of each participant’s edited and distilled initial interview (the resulting “film clip”), it is shown to each of the other participants. Participants view each others’ clips individually, reflect on their meaning and implications, and frame their questions, comments and responses, which they later share with us on film. This ‘responsive interview’ session is also filmed, analysed, and then reviewed and edited with the participant’s input, until they feel comfortable showing the material to the other participants. This process is repeated a second time and a ‘second clip’ is assembled.
Finally, we review all approved footage and related transcripts. We go back and forth between the transcripts and the film clips. This allows us to examine the important narrative elements of the participants’ contributions textually within the transcripts and to analyse the degree to which the narrative lines we have identified as the core axes of the emerging dialogue actually ‘work’ on film. Subsequent analysis meetings involve sharing our views on how best to represent the dialogue based on the available data. We view preliminary assemblies of all the participants’ film clips and make decisions about which clips and which particular sequences contribute most to the clarity and coherence of the dialogue, and to the visual flow and pacing of the film, according to our analytic criteria (see below).
The ‘first cut’ assembles the final selections of clips from all the participants into a single dialogue. This cut of the film is shown to all participants individually for their review and approval. Participants offer further editorial feedback and can suggest edits to their own footage to ensure that they are comfortable with how they are portrayed in the resulting film. We incorporate this final round of feedback and make any necessary changes, taking care to preserve the integrity and meaning of the dialogue. This process can take several iterations (cut 2), and can require additional filming in some cases in order to clarify specific points or to fill logical gaps in the dialogue. Once all participants are satisfied with the final assembled film—the final cut or cut 3—we seek their permission to disseminate the film. Depending on the context and the sensitivity of the topic, we may seek permission for unrestricted dissemination, or for each specific instance of dissemination. Brokered Dialogue participants have been extremely supportive of this level of control over the representation of their own views, though it raises challenges in terms of the efficiency of production for the final films.
Brokered Dialogue is a qualitative research method with four specific analytic goals: 1) where do participants start? The goal is to understand differences and similarities in perspectives that participants bring to a given topic, i.e. their ‘baseline’ positions; 2) where do they diverge? The goal is to identify points of issue or contention within a chosen topic that can serve as moments of intersection among the participants’ perspectives and therefore also as key elements upon which to build a dialogue; 3) how do they change? The goal is to look for signs of evolution or change in individual participants’ perspectives that could be attributable to participating in the dialogue; and 4) what are the pathways to progress? The goal is to identify promising or novel pathways for progress towards solution or resolution of the issue under consideration that are revealed through dialogue.
Interviews are videotaped and audio taped to allow us to capture interactions between researchers and participants that occur off camera and incorporate them into the analysis. Audio tapes of all interviews are transcribed verbatim. We work from the text-based data to document emerging themes initially and to inform the editing process. We review the developing iterations of the film footage and use both written transcripts and the visual footage to inform the final rounds of edits.
Transcripts are reviewed and coded independently by all members of the project team. The dialogic narrative analysis is reviewed at team meetings and consensus reached. Editing decisions also incorporate the emerging analysis in order to organize the raw footage thematically as the resulting film is constructed. Accepted techniques for ensuring analytic rigour and trustworthiness for qualitative studies are employed, adopting a ‘relativist’ approach as characterized by Sparkes and Smith (2009) and others [3
Narrative analysis of textual data
We use narrative analysis to identify common themes and patterns in the written transcripts [31
] and employ a combined approach of thematic, interactive and dialogic narrative strategies [27
]. This approach analyses “how protagonists interpret things” [33, p. 5] and is concerned not only with the content of what participants say, but the form and tone of their accounts as well [5
]. The experiences and perspectives offered by participants in the Brokered Dialogue
(the protagonists) reveal important aspects of social life [34
]. The interactive narrative approach [27
] is particularly useful for studying the interactions of these perspectives, which is central to the Brokered Dialogue
method. In particular, points of uptake and resistance to others’ perspectives are highlighted as potential moments of intersection of perspectives. A strategy of multiple readings of each transcript is adopted [35
], with some readings undertaken to analyse the transcript as a story in its entirety, and others entailing line-by-line coding [32
]. Constant comparison within and between transcripts is employed [38
]. The embodied, emotional work of this kind of analysis is recognized [3
In keeping with the narrative and dialogic narrative underpinnings of the method, we believe that judgments of inquiry quality should be grounded in Frank’s framework (2012), with its commitment to dialogic rather than monologic design (seeking multiple accounts from numerous tellers), a recognition of polyphony (interaction of each teller’s voice with specific others, both heard and anticipated), a recognition of the differing narrative resources at the disposal of differently positioned participants, and a recognition of the fluid and ongoing evolution of stories offered [2
]. Along the same lines, Brokered Dialogue
analysis is less about summarizing ‘findings’, and instead “aims at increasing people’s possibilities for hearing themselves and others” [35
]. Prolonged and intensive engagement with participants is central and allows the researchers to interact with them and determine the degree to which they are comfortable with their accounts as represented on film and to gauge how their perspectives may have shifted over time. Quality is also assessed based on the coherence of the emerging film, the insights generated, and addressing the narrative tensions that emerge [40
].We do employ techniques of constant comparative analysis, using multiple analysts, record keeping (of both data collection and analytic decisions), but always employing a narrative and non-finalizing lens throughout the project [35
Analytic approach to film editing
Analytic sessions focused on film editing are conducted in addition to the extensive analysis of textual data to determine which thematically-based segments should be incorporated into the final Brokered Dialogue films. Considerations of flow, pacing, timing and length constraints, visual and sound quality, and relative contribution to the resulting meta-narrative are incorporated into the crafting of the final film. The process is analogous to the writing up of qualitative research findings, marshalling evidence for themes through example quotes from participants. However the interaction of participants’ perspectives with one another is also portrayed in order to demonstrate the interests at stake in the topic, the resistance and uptake of other perspectives by particular participants.
The ‘brokering’ process is embedded within the method, as described above. However the core aspect of brokering is analytic – it is the analysts’ role to identify ‘points of intersection’ between particular participants’ accounts which contribute to ‘core axes’ for the dialogue. ‘Points of intersection’ occur within participants’ accounts, where they identify similarities between their perspective and those of other participants, or may emphasize a perspective differing significantly from those of other participants. Evidence of convergence and/or divergence of perspective are important to understanding the controversy and for illuminating possible pathways to solutions or resolutions. As analysts we then bring these convergences/divergences together on screen to make the key insights and opportunities for further dialogue explicit for the viewer. One further brokering role fulfilled by the researchers is to ensure respectful exchanges between participants, and to probe/challenge their positions on the topic at hand, in order to enhance clarity and to improve the efficiency of the dialogue.
Research ethics considerations
Our Brokered Dialogue projects involve human research participants and therefore undergo ethics review by the St. Michael’s Hospital Research Ethics Board. The research ethics implications of conducting filmed interviews are considerable. All participants give written informed consent. All those invited to participate are told repeatedly that the nature of film means that they cannot be anonymous, and must agree to being filmed in order to participate. Consent is seen as an ongoing process, not a one-time event. Because they are recognizable on film, participants are reminded repeatedly that they have considerable editorial control over what gets shown to others. We use a participatory editing process. Because our goal is to promote in-depth and respectful interaction between key stakeholders with varying perspectives on a topic, we emphasize that they are not obliged to share any filmed segments that they would feel uncomfortable showing another person. They are invited to edit out any segments that they would not wish another person to see, or that they worry might be construed as offensive to other interested parties. No material is viewed by anyone other than members of the research team without the participants’ input and approval. Unlike documentary filmmaking, concerns of respect and abiding by participants’ wishes override those related to film quality. This core feature of Brokered Dialogue has been extremely well received by our participants.
While the method is innovative and has numerous strengths, it is not without its challenges. This technique presupposes a familiarity with qualitative narrative analysis and dialogic and interactive approaches to narrative analysis specifically [2
] . For those less familiar with such approaches it may require a considerable investment of time and intellectual energy. In addition there are technical challenges inherent in working with filmmaking equipment, film editing, and visual data and analysis. We work extensively with experienced documentary filmmakers who are skilled in editing, directing, and camera work. We have found such interdisciplinary collaboration to be very rewarding. Finally, the technique itself, and particularly the participant-driven editing process can prove time consuming and labour-intensive. We intend to explore options for improving efficiency in future work.