Abstract
Personal narrative is at the heart of how human beings share information, represent identity, and convey ideas. Narrative research is a form of qualitative analysis that assists researchers in gaining insight into the lived experiences of the persons they are studying within their unique life circumstances and contexts. Although many narrative investigations report themes from study data, there is no single, well-defined approach to data analysis in narrative research. In this article, we provide a method for analyzing the data beyond the spoken words by applying Riessman’s thematic, structural, and performance analysis. We describe how applying multiple methods of systematic evaluation to narrative data leads to a deeper and more valid insight into the told stories. The data analysis process outlined in this article contributes to the academic discourse and knowledge supporting the use of multiple methods of systematic evaluation to uncover deeper meaning and thus leading to greater validity of the findings from narrative data.
Keywords
Introduction
In recent years, the narrative research approach has become an increasingly valued qualitative methodology for exploring and appreciating human experience among researchers from multiple disciplines such as psychology, history, sociology, and philosophy (Riessman, 2012). Similarly, there is a growing interest in utilizing narratives in nursing research to understand, from the patient’s perspective, the experience of living with illness (Holloway, 2007). When patients are asked to describe disruptive life events, such as managing a chronic illness, many respond with stories (Riessman, 2012). Patients articulate their personal story from a point of view that is not just an account of facts but instead a narrated reconstruction of how they interpret and understand their past, present, and even future health.
Despite a substantial interest in narrative research in nursing, methods of analyzing patient stories for clinically relevant meaning are still being developed and are infrequently discussed (Casey, Proudfoot, & Corbally, 2016; Lewis, 2018). Much of the literature discusses the benefits of narrative research methodology for providing insight into peoples’ health issues (Casey et al., 2016), yet very little is written about rigorous methods of narrative analysis to generate meaningful and useful results (Maher, Hadfield, Hutchings, & de Eyto, 2018). As narrative research becomes more prevalent in nursing research, it is imperative that narrative data analysis is conducted in a rigorous and methodical manner to yield valid and reliable evidence to be utilized in practice. We argue that applying multiple analytic methods, as suggested by Riessman (2008), to interpret narrative data, leads to deeper and more valid insight.
The application of multiple methods of systematic evaluation to interpret narrative data is discussed in this article. While there are many ways to conduct narrative research and analyze the data, this article will focus on Riessman’s (2008) approach to narrative analysis and three particular methods of interpreting and systematically evaluating told stories. These methods are thematic, structural, and production analysis. Using an exemplar study investigating older adults’ decisions for medication nonadherence, this article will describe our experience with locating deeper and more valid insight using Riessman’s narrative analysis.
Validity in Narrative Research
The concept of validity has been established as a tool to test the quality of research claims in terms of their likely truth. According to Polkinghorne (2007), the concept of validity concerns the believability of a study finding or knowledge claim. The judgment about the validity of knowledge claims is based on the weight of the evidence and the argument offered in support. Thus, a conclusion is considered valid when there is sufficient evidence to reasonability believe it is true (Hammersley, 1990).
Many qualitative researchers argue that assessing validity is no longer useful since the meaning of this concept was originally based on realist or positivist assumptions, making it inappropriate for naturalistic forms of inquiry (Lincoln & Guba, 1985; Rolfe, 2006). Others reason that rejecting realism for relativism as the basis for evaluating research may devalue the seriousness of rigor in qualitative research (Maxwell, 1992; Porter, 2007). Although there are opposing schools of thought about assessing validity in qualitative research, it is essential to assess the likelihood that a claim is true before it can used to inform nursing practice (Porter, 2007). Whether using the term credibility, explanatory power, compellingness, or trusthworthiness, the validity of research claims is an important aspect of conducting quality research in any paradigm (Hammersley, 1990; Heikkinen, Huttunen, & Syrjala, 2007; Leung, 2015; Lincoln & Guba, 1985; Noble & Smith, 2015; Porter, 2007; Sandelowski, 1991).
Maxwell (1992) suggested that one of the most important types of validity in qualitative research is interpretive validity. This type of validity refers to the evaluation of the participants’ perspective and is particularly concerned with constructing accurate meaning from participants’ accounts (Maxwell, 1992). We suggest that robust data analysis procedures can be a way for bolstering confidence in the degree to which a story accurately represents the phenomena being studied. The potential for strengthening the validity of the research findings is the motivation for applying multiple methods of systematic evaluation in the analytic process.
Riessman’s Narrative Analysis
Important to narrative research is the assumption that meaning is constructed from personal stories (Polkinghorne, 1988; Riessman, 1993). Through storytelling, individuals recapitulate and reinterpret their lives for a listener. There is choice and individuality in how experiences are perceived and shared. When recounting events, individuals actively organize a story to tell about past action and how they understand those actions (Riessman, 1993). According to Riessman (1993, 2008), events become meaningful because the individual chose to include them in the story. The story itself, however, goes beyond a retelling of events. The story exposes the audience to the experience from the teller’s point of view, revealing intentions, motivations, decisions, thoughts, emotions, actions, and consequences. The stories provide models of the interrelationship between actions, thoughts, feelings, and consequences (Riessman, 1993, 2008).
Attention to narrative as both an approach and an object for analysis is a defining characteristic of Riessman’s (1993) narrative analysis. Since personal stories are believed to be a window into the individual (Riessman, 2008), narrative analysis has the power to construct meaning from people’s told stories. Narrative analysis has no unified rules about methods for evaluating the data (Riessman, 2008). Narrative analysis has been described as a “cluster” (Figgou & Palopoulos, 2015) of analytic methods and a “methodological repertoire” (Quinn, 2005, p. 6). A good narrative analysis takes the reader beyond the surface of a text. The story should not be fragmented or reduced to mere themes but rather held as a discrete unit that can be systematically evaluated for deeper meaning (Riessman, 1993, 2008). Data analysis for qualitative narrative research tends to be complex; therefore, some narrative analysts believe the application of multiple methods of evaluation is critical to elicit the most meaning from the data (Casey et al., 2016; Maher et al., 2018; Riessman, 2012).
Riessman’s (2008, 2012) narrative analysis acknowledges four general analytic approaches, but for the purposes of this article, only three will be presented: (1) thematic, (2) structural, and (3) performance. Systematically interrogating the narrative data with these different methods of evaluation is meant to take the analysis beyond a surface reading of the text and uncover the underlying meaning embedded in peoples’ stories. Riessman (2012) allows freedom of the researcher to apply one or all of the analytic methods as needed to elicit the most meaning from the story. Locating deeper meaning in a narrative story requires going beyond what was said to include analyzing how and to whom the story was told, through application of multiple methods of systemic evaluation to gain deeper understanding of the told stories. Through this critical evaluation approach, the researcher will discover the deepest meaning and interpretation (Casey et al., 2016).
The following narrative study is utilized as an example to illustrate the application of multiple methods of systematic evaluation in analyzing stories and the increasing of validity during the process. The study is briefly described.
Exemplar Study
The exemplar study utilized Riessman’s framework for narrative analysis to explore the medication-taking decisions of older adults with heart failure (Meraz, 2017). The study aimed to understand the role of decision-making in medication nonadherence, how older adults make decisions for medication nonadherence, and what older adults consider medication adherence and nonadherence.
This qualitative investigation was conducted by the lead author in a large metropolitan area in the Southwestern United States. Since an important aspect of generating quality study data is dependent upon the study participants having a story to tell (Riessman, 2008), sampling was purposeful, seeking to recruit older community-dwelling adults with medication-taking stories to tell. Eleven older adults between the ages of 69 and 92 (M = 80.7 years) participated in the study. The study was institutional review board approved, and all participants completed informed consent prior to enrollment. The qualitative data collection phase consisted of in-depth, semistructured interviewing that was conversational in nature. Interviews were audio-recorded and then transcribed for analysis. The study data from each interview was systematically evaluated using three analytic methods, thematic, structural, and performance analysis. QSR International’s NVivo Version 11 qualitative data analysis software was used to support the qualitative analysis and organize narrative data.
Thematic Analysis
In the thematic analysis approach, content is the exclusive focus, giving primary attention to what was said. Thematic analysis asks, “What was the point to the story?” “What were the told events?” and “What main idea or theme is directly or indirectly stated?” (Riessman, 2008, 2012). Different from other qualitative methods, such as grounded theory, Riessman does not offer a set of rules for analysis. In Riessman’s thematic analysis, the story is kept intact and not fragmented into codes or segments. Themes developed by the investigator may retain multiple influences including prior theory, the aims of the investigation, and previous knowledge. Researchers immerse themselves, through numerous interactions with the data, to familiarize themselves with the content of each story.
Exemplar—Thematic analysis
In this study, locating themes began during the interview process. Participants were asked to tell their story about a time they chose to take a heart medication differently than prescribed. Listening and interpreting what was being said informed the formation of follow-up questions. The interviewer documented thoughts, insights, and observations and noted emerging patterns in subject matter.
Transcribing the narrative data is the next interpretive phase of thematic analysis. Included in the systematic evaluation of our study was the application of multiple transcription techniques. First, interviews were transcribed verbatim to include pauses, nonlexicals, and emotions of how the narrative was delivered such as laughing, crying, and tone of voice. Observations from the field notes, such as the setting, were noted in the transcript. Since interviews were conversational in nature, talk unrelated to the study topic was common and the first version of the interview transcription was long. After multiple readings, the decision was made to minimize (the text is not removed but reduced to a very small font size of 6) conversation unrelated to the study topic, reducing the transcript to the essential back-and-forth discourse between the participant and the researcher. The decision was made to minimize the text to a small font instead of removing it so it could easily be reversed. Table 1 is an example of a transcribed story about stopping a medication and the unrelated conversation removed. Drawing from Riessman’s approach to analyzing the intact story, the transcripts were pared down further, removing the interviewer’s probes for more information (i.e., “Tell me more about that” and “What happened next?”), documenting an uninterrupted and intact story. Table 2 exemplifies the intact story.
Transcribed Story With Unrelated Conversation Removed.
Transcribed Story Intact.
Thematic analysis was used to interrogate participants’ stories for meaning about medication-taking decisions. As a starting point to exploring the participants’ experiences, the overarching question was asked, “What was the point to the story?” The focus for analysis was each participant’s whole story, resisting the tendency to fragment the story or compare themes with other data. Looking through the lens of a theoretical framework, the researcher ventured deeper in analysis. The theoretical framework added a rich dynamic to uncovering the motivations and emotions embedded in participants’ stories. Although each story was thematically analyzed separately, many themes were shared across the narratives. We believe this strengthened the interpretive validity of the investigation.
Structural Analysis
The structural analysis approach analyzes the structure of the told story. The classic story structure has an introduction and plot, along with characters, a problem or conflict, climax, resolution, and coda. Structural analysis directs the researcher’s attention to how the story was told, organized, and put together to depict the intended message. This analytic approach shifts from examining “What happened” to how the participant shaped their story, attempting to persuade the listener to his or her point of view. Like thematic analysis, structural analysis is concerned with content, but also queries, “How was the story constructed?” “Are there any gaps or inconsistencies in the story?” “What details (such as characters) did the teller choose to include in the story?” and “What is the sequence of events?” (Riessman, 2008).
Exemplar—Structural analysis
For this study, structural analysis was the second method of systematic evaluation. Compared to the broader lens of thematic analysis, structural analysis focused on the details and particulars of the story and story structure. To facilitate structural analysis, participant stories were arranged in chronological order by story element which generated the development of plot diagrams. A plot diagram is a literary tool used to analyze and understand written work by visually organizing a story by introduction, rising action, climax, falling action, and resolution (Dobson, Michura, Ruecker, Brown, & Rodriquez, 2011). Plot diagrams containing the classic story elements of setting, introduction, plot, a problem or conflict, climax, resolution, coda, and characters were developed for each study participant. Each aspect of the story structure became evident through this form of transcription and analysis. An example plot diagram is presented in Table 3.
Example Plot Diagram.
Structural analysis was a stronger analytical technique than anticipated. It was expected that the analysis would expose gaps in story structure and present the story in chronological order. Structural analysis, however, was also effective in highlighting what drives actions. The researcher conceived that the participant stories had the power to reveal intentions, motivations, decisions, thoughts, emotions, actions, and consequences and found structural analysis beneficial in locating interrelationships between these concepts. For example, the analysis found the problem or conflict in the story to be a motivator to choosing to take medications differently than prescribed. The climax of the story indicated there was a relationship between thoughts about a medication and medication-taking actions, while the resolution revealed the consequences of the medication-taking decision. The coda, or concluding remark, revealed the participant’s perception of finality to the medication-taking decision. Because structural analysis identified the setting and characters in participants’ stories, this form of analysis also illuminated the role of external influences on medication-taking decisions.
Locating and evaluating the structural elements of participant stories pushed the analysis beyond what was said to examine how the participant chose to formulate their story, revealing new findings. In particular, this analytic approach illuminated participants’ decision-making process in a way that thematic analysis could not. When the structure of the story was diagramed in chronological order, a consistent sequence of events emerged from the stories. This analytic approach was a powerful tool for constructing a model explaining the process older adults use to make a medication-taking decision. In sum, reinforcing thematic analysis with structural analysis helped to clarify ambiguities, illuminated insights that might have been missed, and aided in determining the relationship between feelings, thoughts, and actions, and uncovered new findings.
Performance Analysis
Performance analysis asks to whom the story is directed and for what purpose. Without a listener, there is no story. Therefore, the teller constructs the story for a particular listener with language, vocal tones, and gestures that convey how they want to be perceived. In situations of potential stigma, Riessman (2012) asserts that storytellers perform “a more desirable self” to protect their true identity or actions from the listener. Performance analysis understands that storytellers will work to pull the listener into their experience and persuade them of their point of view. Questions that can be considered are as follows: “Did the participant make a concerted effort to perform what they think is more desirable to the listener?” and “Why was this particular story performed?”
Exemplar—Performance analysis
Performance analysis was the third analytic method used in the study. This analytic approach was included in the systematic evaluation of stories given that previous research reports that patients may be reluctant to disclose decisions for medication nonadherence to health professionals for fear of judgment (Brown et al., 2016). Performance analysis was important to explaining why participants acted out their story or included specific details. The method for applying performance analysis unfolded as data collection progressed.
The application of the performance analysis approach took two directions. First, field notes and reflexive journaling were important to capturing data about how the story was delivered, such as body language. Some participants in the example study demonstrated a concerted effort to perform or act out a point they were trying to make. For example, one participant intentionally wore clothing stained with blood to convey thoughts and feelings about the bleeding effects resulting from anticlotting medication use. Upon introduction, this participant pulled up her sleeves and pants to show bruises. She pointed out stains on her clothing and acted distressed.
The second application of this analytic method was through multiple readings of the interview transcripts, plot diagrams, and immersion in the stories. Upon interrogating the stories, it was noted that retired participants were consistently including in their stories a past profession or job. Examining the pattern with a performance analytic approach illuminated the participants’ tendency to include past professions as examples of their competent decision-making ability. Performance analysis revealed that when participants referred to a job or wore clothing stained with blood, they were forming an argument to justify their medication decisions and convince the interviewer they were not nonadherent but rather good decision makers that were making necessary decisions for their well-being. The analytic approach of performance analysis added a rich dynamic to data analysis, giving way to clarifying and finding new themes.
Validity and the Application of Multiple Analytic Methods
We propose that applying multiple analytic methods to interpret narrative data strengthens the validity of the findings. The judgment about the validity of study findings is based on the weight of the evidence and the likelihood that a claim is true (Hammersley, 1990; Polkinghorne, 2007). Although knowledge gained from qualitative investigations have the potential to inform nursing practice, the validity of the claims should be established before translated to care. We suggest that robust data analysis procedures, as described in this article, can increase confidence in the degree to which a story accurately represents the phenomena being studied. Through the application of three analytic approaches: (1) thematic, (2) structural, and (3) performance, narrative analysis located deeper meaning, clarified ambiguities, and increased understanding of the experiences told through the story.
In the exemplar investigation, the three analytic approaches were applied to each participant interview. Thematic analysis was a good starting point for interpreting meaning from participants’ stories. We argue, however, that thematic analysis alone would have limited the interpretative process and the quality of the findings.
Although structured analytical techniques are not the goal of narrative analysis, locating story structure during structural analysis affirmed and clarified thematic analysis findings. It also revealed new findings that thematic analysis alone did not capture. Most importantly, structural analysis was conductive to developing models of medication-taking decisions, drawing relationships between participants’ thoughts, motivations, and actions. This analytic approach took the analysis beyond mere themes. The performance analytic approach further affirmed and clarified thematic findings yet also yielded more unanticipated emergence of themes.
The presentation of study findings is an important aspect of utilizing research to inform nursing practice. Although study data are analyzed utilizing multiple analytic methods, presentation of study findings is not fragmented. The insight gained from each method of data analysis is amalgamated into themes and models that represent the underlying meaning embedded in participants’ stories. Since applying multiple analytic approaches to stories can generate deeper and more valid insight into patient illness experience, findings can inform the development of a more specific individualized plan of care (Casey et al., 2016).
Conclusion
Analysis of narrative data yields the researcher a deeper understanding of the told stories, deriving a deeper meaning from what is told through application of multiple methods of systemic evaluation. The combination of thematic, structural, and performance analysis helps to clarify and confirm study findings and uncover new findings thus strengthening the validity of the findings. The data analysis process outlined in this article contributes to the academic discourse and knowledge supporting the use of multiple methods of systematic evaluation to interpret meaning from qualitative and strengthen the validity of research findings.
Footnotes
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) received no financial support for the research, authorship, and/or publication of this article.
