Methodological background
The construction of composite narratives as a research output has been employed by researchers using a variety of qualitative approaches and methodologies (including phenomenology, ethnography, narrative research and grounded theory) and a range of methods of data collection such as interviews [
8,
23,
36,
39], written biographies and autoethnography [
8], creative writing and artwork [
37], photographs and field notes [
42,
43], and audio recordings of workshops [
37]. Composite narratives can form one part of the outputs of a study [
11,
18] or be the main focus of the data [
8,
37,
38].
Composite narratives can be used either as a methodological approach integral to the research design, analysis and outputs [
8,
37], or as a vehicle for presentation of research outputs which may be decided on after completion of the initial analysis [
11,
18,
23,
39]. The methodological approach and philosophical underpinnings of the research influence the form of the composite narrative constructed from the data. Most research employing composite narratives is from a constructivist philosophical standpoint, subscribing to the belief that all knowledge is subjective and socially constructed [
40]. This is important because such a philosophical stance acknowledges the importance of the researcher’s interpretation in the research process and therefore accommodates more involvement of the researcher’s voice in the findings than a realist or positivist approach would allow
—such approaches requiring the bracketing or separation of the researcher’s influence on the results.
Examples from the literature of ways in which composite narratives are constructed include first-person narratives [
8,
18,
37,
38,
42,
43], third-person narratives [
11,
23,
39], and ‘composite fictions’ [
24]. Tab.
2 summarises examples of differences in the methodological approach, method of data collection and type of composite narrative produced from a few key studies. The table draws on examples from health and social sciences research, as examples from the medical education literature are limited.
Table 2
Methodological approaches to constructing composite narratives
Ethnography | | Narrative interviews Written narratives (n = 25+) | ‘Composite fictions’ First-person account with characters, settings, context, and plots created to protect identities of participants |
Phenomenology | | Creative writing, drawing, audio recordings of workshop (n = 5) | First-person narrative Three research group members each produced a composite narrative, these were combined to produce a single final narrative |
| Interviews Nosek (n = 19) Wertz (n = 15) Marlow (n = 17) McNeish (n = 15) | First-person narratives Aim to convey both structure (narratives illustrate the main themes) and texture (richness of participants’ experiences) |
Narrative research | McElhinney and Kennedy [ 8] | Interviews, Biographical written narrative, Autoethnography (n = 9) | First-person narratives Combined narrative data from participants with autoethnography, narrative arcs, phraseology and style of individual narratives incorporated |
| Interviews (n = 14) | Third-person narratives Descriptions of interviews with composite characters. Direct quotations from transcripts incorporated into narratives |
| Interviews (life history research) (n = 3) | First-person narrative Combined text from all interview transcripts to create a composite |
Qualitative framework analysis | | Interviews (n = 48) | Third-person narratives Scaffolded by theoretical framework |
Grounded theory | Johnston, Wildy & Shand [ 33] | Interviews (n = 25) | First-person narratives 24 narratives produced, one to illustrate each of the findings |
Researchers in the phenomenological tradition use composite narratives as a way of conveying both the structure (the themes characterising the phenomenon) and the texture (the richness of the experiences of the participants) of the research findings [
18,
34]. This is in keeping with the aim of phenomenological research to describe and explain the subjective lived experiences of individuals [
40], and its emphasis on the importance of the first-person viewpoint in conveying this richness of experience. In healthcare research, phenomenological researchers argue that composite narratives are an accessible means of conveying research findings which can enrich healthcare practitioners’ understanding of their patients’ lived experiences and facilitate empathic care [
18,
37].
Narrative research sets out to explicitly gather data with an emphasis on the story of an aspect of an individual’s life and is a natural environment for the evolution of the composite narrative as a research output [
8,
23,
38,
39,
42,
43]. Narrative research acknowledges that the data gathered illuminates both the participants’ experience and their reflections on the experience as they frame it for telling, choosing language, emphasis and shared reflections. Composite narratives produced through narrative research include both first [
8,
38,
42] and third [
39] person composites. In our own research, we used data in the form of written narratives, interviews and an autoethnographic study to produce composite narratives describing the experiences of academic GPs [
8]. Autoethnographic study was included as one of the authors was an insider researcher and this informed the choice of the first-person voice, emphasising the intertwining of the researcher’s voice with those of the other participants in creating the narratives from their shared experiences. This research is an example of the composite narrative being integral to the design, analysis and outputs of the research; the decision to create composite narratives was embedded in the research design and specified when consenting participants. Written narratives were first analysed individually taking a thematic narrative approach [
5], with similarities of the narrative arcs and shared experiences between groups of participants identified. The authors then started drafting composite narratives for each of these groups, highlighting these commonalities in the narratives. This initial stage of the analysis helped to shape the focus for the subsequent interviews which explored the similarities and differences between the participants in more depth. The texts from the transcribed interviews were then also analysed individually, before being synthesised with the three developing composites. Taber [
38] and Marsh et al. [
42] also created first-person narratives to present their research findings. In both these studies, the authors created their narratives by blending the words of all participants to create a composite.
A different approach was taken by Willis [
39], in her research exploring how senior politicians understand their role in relation to climate change. Following the initial analysis and coding of her data using NVivo software, she found that ‘The overall picture thrown out by the interview data was both more tangled and richer than anticipated’ with a ‘complex web’ of ‘individual motivations and outlooks’ [p. 481]. Her decision to use composite narratives was taken as a means of addressing the challenge of conveying ‘the richness and complexity of individual accounts’ while honouring the ‘need to ensure anonymity’ [p. 481]. She explains that she chose to use composite narratives to protect the identities of her participants and to explore context and complexity rather than defining differences and categories [
23]. Willis developed four composite narratives presented as reports of interviews with a composite character, using quotations derived from the interview transcripts.
In their ethnographic study exploring the experiences of teachers wrongly accused of sexual abuse of pupils, Piper and Sikes describe the creation of ‘composite fictions’ which they describe as ‘bring[ing] the written product of social research closer to the richness and complexity of the lived experience’ ([
45, p. 7–8], cited in [
24, p. 568]). Given the sensitivity of their research subject, they needed to convey the experiences of their participants while avoiding any potential for identification from any details presented in the research findings. They addressed this by creating ‘composite fictions’ where they created settings, context and timelines while ensuring they did not invent the experiences and perceptions of participants conveyed in the narratives ([
36, p. 43–44], cited in [
24, p. 568]).
More recently a large qualitative study exploring the wellbeing of doctors in Ireland during the COVID-19 pandemic [
11] produced composite narratives that summarised their key findings as third-person accounts, scaffolded by the theoretical model they used to analyse their large body of qualitative data.
Establishing trustworthiness
As Polkinghorne [
46, p. 471] notes, ‘attention to the judgments about the validity of research—generated knowledge claims is integral to all social science research’. Some readers of research, particularly those working in a positivist or post-positivist research paradigm, may question the validity of research that involves the ‘storying’ or ‘re-storying’ of data. Such apprehension about the trustworthiness of the interpretation may be exacerbated where stories are combined and/or joined by the voice of the researcher. A core value of research is that its findings are accepted and that it holds the possibility to influence change [
22]. In order to uphold these values, narrative (and composite) research must demonstrate ‘the procedures used to ensure that its methods are reliable and that its findings are valid’ ([
47], cited in [
22, p. 12]). For researchers working within a constructivist paradigm, however, trustworthiness and the extent to which the researchers or relevant communities can feel safe in acting upon the findings [
48] are held to be more helpful quality criteria than validity.
Attention to the elements of trustworthiness (credibility, transferability and dependability) as set out by Lincoln and Guba [
49] are important in establishing the value of qualitative research methods. This is particularly true for methods that are novel within a particular field. With this in mind, medical education researchers creating composite narratives should consider steps such as prolonged engagement with their subjects, triangulation of data, peer debriefing, archiving of raw data for future comparisons and member checking as vital components in establishing credibility [
49]. Transferability in the constructivist context is not directly comparable with the external validity expected in the positivist paradigm. However, the provision of detailed description and contextualisation will allow readers to make informed decisions about the appropriateness of transferring the themes of the findings and acting upon them in their own context. Finally, dependability can be strengthened by having multiple researchers involved in the design, analysis and outputs of the research. Establishing the trustworthiness of research findings presented as composite narratives therefore requires clear description of the research design including the steps taken to establish credibility and of the approach taken to transform individual narratives into composites [
50]. Composites must include thick description and context in order for readers to make decisions about transferability of the findings.
Reflexivity is also vital in ensuring the trustworthiness of this method and the researcher’s role should be acknowledged and accounted for. In some methodological approaches, such as interpretative phenomenology and those that combine autoethnography with other participant data, reflexivity is explicit in the method. In interpretive phenomenological research, the interpretation by the researcher of the phenomenon described is a vital aspect of the research [
18] and the composite conveys both the phenomenon and the researcher’s interpretation of its meaning. Similarly, combining autoethnography with data from other participants makes explicit the author’s entwinement with the subject [
8]. Willis [
23] acknowledges her involvement by situating herself in the narratives as the interviewer narrating her third-person accounts of the interviews and including her observations on the composite characters described.