Abstract
Narrative research methodology is evolving, and we contend that the notion of emergent design is vital if narrative inquiry (NI) is to continue flourishing in generating new knowledge. We situate the discussion within the narrative turn in qualitative research while drawing on experiences of conducting a longitudinal narrative study. The philosophical tensions encountered are described, as our understanding and application of narrative approaches evolved. We outline challenges in data collection and analysis in response to what we were learning and identify institutional barriers within ethics review processes that potentially impede emergent approaches. We conclude that researchers using NI can, and must, pursue unanticipated methodological changes when in the midst of conducting the inquiry. Understanding the benefits and institutional barriers to emergent aspects of design is discussed in this ever-maturing approach to qualitative research.
Keywords
Our experience in launching a narrative study reinforced the emergent, nonlinear, and often messy nature of qualitative inquiry. It also foregrounded how institutional ethics review boards are increasingly calling for predetermined, step-by-step outlines for research projects that can unwittingly constrain emergent approaches. We draw on experiences from a longitudinal narrative study, Re-stor(y)ing Life Within Life-Threatening Illness, to describe how our thinking and research processes evolved and diverged from the “research as planned” once we began collecting data with people living with serious illness. Rather than interpreting such divergences as problematic, we explore the importance of emerging designs in order to advance narrative research and sustain meaningful knowledge development.
We situate this discussion in concepts of the narrative turn and emergent design in qualitative research. Next, we introduce a narrative inquiry (NI) that informs this discussion and describe our evolving circular approach. This includes expanding our understanding of narrative methodology (and philosophical tensions), data analysis, and data collection that led to unexpected forms of analysis, knowledge generation, and innovative dissemination. In addition, institutional barriers to emergent design from increasingly detailed research ethics review processes are addressed. Key strategies for embracing the unexpected in NI and institutional constraints that minimize emergent aspects of NI are also explored.
Narrative Turn
The narrative turn (Polkinghorne cited in Goodson & Gill, 2011) is a term used primarily in literary studies, social, and human sciences and expresses a shift toward legitimizing peoples’ stories as important sources of empirical knowledge (Hyvarinen, 2010). Although it is difficult to articulate an exact time frame, the turn toward narrative can be situated within the “science wars” of recent decades (Gergen, 2015). Challenges to previously dominant postpositivist approaches have been well documented by scholars using qualitative and interpretative methodologies such as hermeneutics, phenomenology, critical feminist approaches, institutional ethnography, and discourse analyses to name a few (Munhall, 2012; Racine & Petrucka, 2011; Wertz, 2011). For the past 25 years, health researchers have published rigorous qualitative research in journals such as Qualitative Health Research and Qualitative Inquiry. It is tempting to assume that contentious paradigmatic debates of “qualitative versus quantitative” are historic; however, deep-rooted beliefs about knowledge and knowledge generation are closely tied to disciplinary values. The growth in funded research, dedicated journals, and conferences reflects a widespread acknowledgement and integration of qualitative methodologies across disciplines; nevertheless, diverse disciplinary values remain and are brought to bear in reviewing funding and ethics applications. Our discussion of emergence calls for research designs that are open ended and by necessity fluid and less structured; traditionally, such emergent approaches have been marginalized in favor of deductive strategies. Even so, the narrative turn toward peoples’ narratives and their expertise regarding their health and illness experience continues. Within the broader domain of qualitative research, NI which is our methodology of interest remains a developing field. For the purpose of this article, we use “NI” to mean the study of experience as story and a way of thinking—through storying (Clandinin & Connelly, 2000).
Emergent Design
Emergent designs are a hallmark of qualitative research and distinguish qualitative approaches from quantitative methodologies. Morgan (2008) describes emergence as a circular process. As new data are being collected, there is ongoing analysis, “so that both the research procedures and questions can be adjusted in an iterative fashion in response to what is being learned in the field” (p. 246). The notion of emergence can range from evolving data collection to unexpected procedures for analyses that change as the findings and new knowledge are generated.
This flexible approach contrasts quantitative approaches, where research protocols privilege operationally defined variables, concise data collection points, and predetermined analytic procedures. The exactness of quantitative designs supports deductive reasoning and analyses. In contrast, qualitative approaches generate knowledge using inductive reasoning, starting with what participants say or enact (raw data) while researchers theorize from grounded data rather than applying apriori theories. While few studies use completely open-ended designs or would be considered atheoretical, emergent aspects are inherent in most qualitative projects (Creswell, 2015; Morgan, 2008).
In a recent NI, we anticipated and planned for some degree of emergence especially when conceptualizing longitudinal interviews; however, the high degree of emergence we encountered could not have been foreseen and is explained in detail below. We acknowledge the richness gained from research that is responsive to what is being learned in the field and also the challenges and costs of needing to change consent forms, revise ethics applications, and wait for institutional approvals when emerging directions deviate from the initially proposed research.
Longitudinal Qualitative Inquiry
The genesis of our evolving understanding of narrative methodology was a longitudinal NI entitled Re-stor(y)ing Life Within Life-Threatening Illness (Sheilds et al., 2015). The purpose of the study was to understand experiences of living with and restor(y)ing life when faced with the uncertainty of life-threatening illness. Over the study’s 5-year period, our team of four researchers and three doctoral students explored related concepts such as liminality, secrets, and spirituality. A qualitative design using NI was planned. We drew on Riessman’s (2008) narrative methodology and Gergen’s (2004) social constructionist perspective to frame the study. The research was further underpinned by four commonly accepted concepts in NI: meaning making (the belief that stories hold meaning; Frank, 2009; Josselson, 2011; Kleinman, 1988; Riessman, 2008), close study of the particular within individual stories as a means to illuminate universals in human experience (Charon, 2006; Riessman, 2008), social constructionism (meaning is cocreated and coconstructed; Crotty, 1998), and the role of metaphoric language and metonymies in stories as a way to understand and talk about illness experience and as a way to understand how participants can hold multiple seemingly contradictory stories at one time (Aoki, Pinar, & Irwin, 2005; Frank, 2011).
Participants
Thirty-two adults living with cancer, end-stage renal disease (ESRD), or HIV/AIDS agreed to participate. Participants were interviewed in a conversational, semistructured format, 4 times each over a 3-year period using a relational, discursive approach aimed at generating detailed accounts. During the second interview, participants were asked to identify a symbol (e.g., photograph, object) that represented their experience of living with serious illness.
Analysis
We proposed both thematic analysis and a case-centered approach to emphasize individual agency, particularities, and context (Riessman, 2008). We first read transcripts as individual stories and then across illness groups. Collaboration by the research team was integral to the evolving ideas about NI that follow. We debriefed after interviews, participated in full-day analysis retreats as a team, held regular meetings wrote group papers, and copresented findings at conferences. Connections established through the combination of these collaborative processes that were frequent, in depth, and nurtured over several years helped to foster dynamic discussions that were exciting and synergistic and rich opportunities to challenge ideas. It was in this iterative process of engagement with participants’ stories, one another, and the literature that our research process began to evolve in unexpected ways.
Evolving Circular Approach
What follows are reflections on how our understanding of NI was (and is) evolving. The methodological unfolding is messy and circuitous and continues to develop through deepening understandings and tensions that underpin the Re-stor(y)ing study. These new ways of thinking led to unanticipated data collection and data analysis.
Evolving philosophical understandings
In designing the Re-stor(y)ing study, we outlined our goals, objectives, and proposed lines of inquiry in order to meet the established goals of the study. We generated a rationale for the study based on a logical integration of concepts. While most of the proposed activities unfolded as planned, things began to change as we immersed ourselves in participants’ stories and their symbolic representations over the 5-year study period. Our assumptions of foundational concepts including what constitutes a story and narrative, for example, were challenged as we encountered people’s expressions that were detailed, coherent, and clear alongside other narrative transcripts that were seemingly incoherent, multilayered, and fragmented. This questioning by coinvestigators and doctoral students added a rich dynamic to the research process and alerted us to the disconnects and shifts in our thinking about ontological assumptions, such as the nature of stories and storying in NI. A number of philosophical tensions had implications for data collection and analysis. We identify four key tensions that came together, creating a force and energy that kept pushing our analysis and knowledge translation ideas toward broader and more creative directions than we had imagined.
The first tension was “What is distinct to NI and narrative analysis?” As our data pulled us in many directions (voluminous interview text, field notes, symbolic representations, narrative descriptions of symbols), for analytic direction, we returned to our understanding of NI and the scholars we were drawing on (Frank, 2002, 2010; Reissman, 2008). Our foci were on stories of individuals and their experience of living with cancer, ESRD, or HIV/AIDS. We took into account intersections between the personal and contextual worlds, peering through a social constructionist lens toward narrative to examine and explore experience and meaning of serious illness in people’s lives. Revisiting our analytic approach, we affirmed that our analysis was about examining stories: seeking to understand and interpret, focusing on the particular, and using the story as the anchor of analysis.
We accepted that the analysis could take a variety of forms, and it could be thematic or relate to a concept or phenomenon of interest such as spirituality or secrets. That analysis could involve whole stories, a participant’s narrative context, stories within stories, and particular exemplars. This remained relevant, and yet it was still not sufficient to hold what we wanted to address in the data, that is, aspects of construction, coconstruction, fluidity, expansiveness, and creativity. We felt a need to add a combination of analytic approaches, and these included using interviewer analyses (reflections, memos, debriefs), team analyses (single interview of a participant, then set of interviews for a participant, then all interviews for an illness group, then across all illness groups), coding analyses, and visual and audio analysis. While confident we held true to traditional understandings of what makes NI and narrative analysis unique, what shifted was where we looked to inform our work. Philosophers and narrativists cited in the original research proposal still informed our work, but we began to seek out, lean toward, and draw in additional narrative thinking that could hold our evolving ideas without feeling narrow or prescriptive. For example, we took into account ideas such as third space and expanding narrative approaches such as that of Pellico and Chinn (2007).
A second tension was a key query: “How do we each continue to understand and differentiate between what is a narrative and what is a story?” In the original proposal, no explicit distinction was made between story and narrative, and the terms were seen to reflect one another and used interchangeably as is common practice for many narrative authors. From immersion in the data, and literature and emerging concepts that called the distinctions into question, the research team shifted; the interchangeable use of the terms at times had us stumbling over them. We questioned whether distinguishing would be helpful. In-depth discussion had us arrive at a shared understanding of narrative as the more formal, broader concept holding all discourses (written, spoken, evoked by image) and stories as smaller expressions contained with(in) narratives. We conceived stories as rooted in lay language, defined by particulars such as emplotment (Polkinghorne, 1991; Ricoeur, 1988) and characters, and having some structure and something happening that is beyond mere description. This emerging understanding shaped coding and interpretive decisions in particular ways.
A third tension was “What is the difference between a theme and a story line
And finally, having adopted a view toward story lines, we found ourselves asking “What is the relationship between story lines and metanarratives?” There was unanimity among team members that one informs the other and both story lines and metanarratives are socially shaped. As the prefix “meta” denotes, we determined that metanarratives will be those overarching or higher order, grand stories that hold social values and power and act as “truths” at a certain time. Whereas story lines are grounded in participant data and come about through narration of an experience, metanarratives lived outside a person and were socially created, becoming so inscribed on a culture and on individuals of a shared society that they shaped individuals’ experiences and thus their stories and the story lines within their more local context. For example, the story line of “preparing for death while living life” and the metanarrative of “it could be worse” can be at play without a person even being aware. Some social metanarratives such as “stigmatizing attitudes” are so taken for granted and become normalized, hardly recognizable, and accepted as the one true story (Beuthin, Sheilds, & Bruce, 2014; Molzahn et al., 2012; Sheilds et al., 2015). We recognized how metanarratives can act as anchors for experience, something to help with sense making and something to hold. What shifted was the tension we experienced related to a new line of analysis; the desire to keep alert to and tease apart the impact of metanarratives on personal stories of illness experience. We read transcripts with close attention to make visible metanarratives that were inferred or implied. We questioned where story lines ended and metanarratives began, and the flow of the influence, always wanting to stay open to the more subtle influences of story lines that come together to support or counter an existing metanarrative or birth a new one. And we continue to question what, if any, particular metanarratives we want to foreground such that they may guide our methodology, somewhat like a master metaphor, and help move us toward concept development.
In summary, the impact of shifting understandings that evolved over time resulted in new ways of thinking about storying, story lines, themes, and metanarratives in our data. Tensions that arose and our ensuing shifts in thinking and analysis could not be known—as is expected practice—when developing the original research proposal. In retrospect, we were surprised by the range of core concepts requiring adjustment and deeper discussions as our analysis unfolded. Even as we write this, we continue to pursue clarity of evolving methodological ideas of what constitutes a story that can be coded as a whole. We hope that sharing our process may resonate with other narrative researchers, spark dialogue, invite curious ponderings, and move narrative into a next turn.
Evolving data analysis and data collection
In addition to surprising philosophical questions, unexpected procedures for analysis were particularly evident during the combined analysis and knowledge translation activities. For example, in the original research proposal, we had planned to develop a full length DVD of living with life-threatening illness. What we did not anticipate was how the participants’ stories would compel us into alternative forms of analysis (e.g., found poetry) and knowledge translation (arts-based approaches) and how the space between analysis and knowledge translation provided opportunities for seeing participants and data in new and different ways (Bruce et al., 2013). For example, in reference to the data collection process stated earlier, we engaged in in-depth conversations with participants on multiple occasions (most often four). During one of those audiotaped conversations, participants were asked to share a symbol that represented their experience of living with life-threatening illness. We then photographed the symbols. What we had not anticipated was the diversity of symbols and range of mediums (e.g., music, hopes, memories), where symbols were of such a personal nature (e.g., an embroidered name on a pillow, picture of family/friends) that further permission from participants for disclosure was sought to address issues of confidentiality.
Further, a decision emerged to replace the DVD plan with an evocative open-access website displaying vignettes using professionally photographed images to covey key findings with voice-over participant narratives. This required collaboration with documentarists and additional taped interviews with select participants. Unexpectedly, when approached, the participants requested their faces and identities to be visible in the photographs and audio website clips (http://www.uvic.ca/hsd/illnessnarratives). In order to incorporate these changes, we needed to submit additional modifications to the ethics review board for approval and engage with participants and family members in numerous conversations.
Institutional Barriers to Emergence
Canadian research agencies and Human Research Ethics Boards (HREBs) often reflect dominant values of certainty, clarity, and control that align with specific empirical approaches to research. Unfortunately, these values may be at odds with qualitative designs that are open, less structured, and therefore better able to adjust research procedures and questions in response to what is being learned. However, while the degree of emergence in a qualitative study will vary, ethics review boards and funding agencies call for specificity, details, and predetermined procedures. Scott and Fonseca (2010) describe experiences, where an ethics committee required, in advance, “all the questions to be put to staff.” This was considered by the researchers to be “an impossibility” (p. 296) because it would render their research ineffectual. They concluded that “the constantly increasing power of ethics committees and their movement towards extreme regimes of regulation” are “overstepping the mark” (p. 298). Perhaps more concerning is an attitude of mistrust that accompanies the increased surveillance by HREBs, when there is little evidence of harm or significant threat to participants (Scott & Fonseca, 2010).
Increasingly, we see cumbersome, time-consuming review processes for relatively minor changes as posing barriers to timely data collection and analysis. By requiring greater detail for ethics approval at the outset, minor changes in procedures necessitate amendment submissions as variations in study participants, contexts, and the “naturalistic environment” are encountered. It seems inevitable that new learning will occur and therefore shape subsequent questions, however methodological decisions can too easily be influenced and thwarted by pragmatic considerations about the need for ongoing institutional approvals. Care must be taken so that the discourses of risk and control which are the purview of institutions do not inadvertently threaten the quality of research and knowledge generated. As professional nurses and researchers, we do not dispute the requirement for ethical oversight by HERBs but see an unintended drift toward greater control and requirements for research designs that minimize sites of emergence and undermine trust of researchers to act ethically. When decisions to not collect additional data or pursue a particular avenue of inquiry are made in order to avoid time-consuming processes of approval amendments, the whole research enterprise is in jeopardy. Our concern is that these forces can constrain the evolution of NI if not kept in check.
Strategies to Foster Emergent Designs
Based on these experiences, we offer four strategies to encourage a return to the importance of emergence in NI. The first strategy is to design research that acknowledges the methodological unknowns by employing open language such as “will be determined,” “in consultation with,” and “pending initial analysis” in proposals. The second is to engage in ongoing discussions among the research team that queries taken-for-granted philosophical assumptions of the analysis being enacted, the phenomena of interest, and how interpretative decisions are being made. Such collective reflexivity builds capacity and breadth in researchers’ interpretive repertoire. That is, ongoing philosophical discussions can extend researchers’ interpretive range “that influences what they see at all, the kinds of research questions they ask, [and] the things taken to be data” (Sandelowski, 2011, p. 344, parenthesis added).
A third strategy is to advocate for pluralistic approaches that encompass methodological synthesis and creativity. Although research textbooks present qualitative methodologies as discrete entities with fixed boundaries, we note that methodologies are often more porous. Such blurring is commonplace in qualitative inquiry, where analytic approaches are routinely intermingled in the actual practice of research (Norlyk & Harder, 2010; Sandelowski, 2011).
And fourth, approaches to enhance trust and understanding between ethics review boards and researchers could be considered. Education sessions can be offered, where needed, to inform HREB members about the aims, processes, and nature of knowledge generated by NI. Researchers can also help to shift the culture of surveillance by volunteering to serve on the HREB, to question and provide feedback where relevant, and to conduct research to better understand how institutional processes are shaping qualitative research. At the policy level, procedures focusing on safety, evidence, and principles rather than rules (Porter-O’Grady, 2010) would better facilitate “right touch regulation,” which is increasingly used by regulators (Bayne, 2012).
Concluding Thoughts and a Call to Action
As qualitative researchers engaged in narrative research, we describe several philosophical tensions experienced when trying to keep our methodology responsive to emerging data. Our aim was to keep our approach open, fluid and evolving, and to stay true to key underpinnings embedded in the concepts of emergence and the narrative turn. We were surprised by the degree to which our thinking and approach began to shift once we entered the nonlinear, often messy, iterative place of data analysis and sense making. Our ability to be responsive and embrace the unexpected was met with barriers in the form of many revisions necessitated by institutional ethics boards. Inadvertently, this privileges more traditional deductive, predetermined empirical approaches to knowledge development. We contend that minor, yet frequent, ethics amendments for changes that are low in risk may thwart creativity and constrain researcher responsiveness to the voices of participants within the very turn intended to shine a light on stories of lived experience. We suggest there is a need to develop sensitivity regarding potential barriers to emergent aspects of NI and advocate minimizing methodological and institutional barriers, when there is limited risk.
In closing, we believe institutional agencies and researchers are not so different, both aim for the highest ethic and examine and hold integrity at the highest level. When we understand the work of ethics boards and researchers this way—as separate but intersecting, as enabling but not restrictive—then we can create a synergy and come together in ways that take us to innovative places. We hope this article sparks debate and turns NI back to its origin of valuing emergence. More broadly, as our understanding has and continues to evolve, we see the critical philosophical tensions outlined in this article as catalytic, guiding us toward a call for evolving qualitative methodologies.
Footnotes
Acknowledgment
We wish to acknowledge the generosity of participants in sharing their experiences of living with serious illness.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study has been funded by the Canadian Institutes of Health Research (CIHR) Grant # MOP-84509.