Journal of Research in Nursing

18(8) 692–704
Keeping the story together: ! The Author(s) 2013
Reprints and permissions:

a holistic approach to sagepub.co.uk/journalsPermissions.nav

DOI: 10.1177/1744987113481781
jrn.sagepub.com
narrative analysis
Claudia C. Beal
Assistant Professor, Louise Herrington School of Nursing, Baylor University, USA

Abstract
Delay seeking medical assistance for ischemic stroke symptoms is common worldwide, yet
existing literature does not provide an adequate understanding of experiences at stroke onset.
An explanatory narrative inquiry using a holistic-content approach to data was conducted to gain
understanding of women’s experiences between symptom onset and hospital arrival. A holistic-
content analysis retains the temporal dimension of each individual’s story so that the parts within
the story are interpreted in relation to other parts of the story and the outcome of the story. The
aim of this paper is to present the procedures for holistic-content analysis, address the
trustworthiness of this form of narrative analysis and appraise the utility and limitations of the
approach. It was concluded that holistic-content analysis is congruent with the philosophical
underpinnings of narrative methodology and it is well suited to examine phenomenon that
have a strong temporal dimension. The use of this approach alone to narrative data does not
enable researchers to draw naturalistic generalisations and a structured comparison of the
narrative accounts resulting from a holistic-content analysis is needed to identify similar and
dissimilar characteristics of participants’ experiences.

Keywords
methods, methodological inquiry, qualitative, stroke, women’s health

Delay seeking medical assistance for ischemic stroke symptoms is an important reason for
patient ineligibility for thrombolytic therapy worldwide (Chang et al., 2004; Deng et al.,
2006; Memis et al., 2008). Most literature on this topic consists of demographic and clinical
correlates of arrival time at the hospital, and there is little explication of the experiential
aspects of stroke onset in the qualitative literature. A fuller understanding of ischemic stroke
onset is needed prior to developing evidence-based interventions to promote earlier help

Corresponding author:
Claudia C. Beal, Louise Herrington School of Nursing, Baylor University, 3700 Worth St., Dallas, TX 75246, USA.
Email: [email protected]

Downloaded from jrn.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on July 11, 2015
Beal 693

seeking for stroke symptoms. This may be particularly important for women who, compared
with men, have greater awareness of the warning signs of stroke (Centers for Disease Control
and Prevention, 2008) but longer delay reaching the hospital after stroke onset (Smith et al.,
2010). Accordingly, the aim of a recent qualitative study on ischemic stroke onset was to
gain understanding of women’s experiences during the period of time between the onset of
symptoms and arrival at the emergency department.
Of particular interest to this investigation was what might account for variations in the
time that women arrived at the hospital after first noticing the symptoms of ischemic stroke.
The temporal dimension of this topic leant itself to narrative inquiry, a diverse qualitative
research methodology in which stories of life events are collected in order to produce a
reconstruction of a person’s experience (Clandinin and Connelly, 2000). Stories are
composed of a unique sequence of events, mental states and happenings that illuminate
the meaning of life events for the narrator (Bruner, 1990). The terms ‘story’ and
‘narrative’ are synonymous (www.merriamwebster.com) and are used interchangeably
throughout this paper.
According to Holloway and Freshwater (2007), the method chosen to analyse narrative
data turns on the issue that formed the basis for the study and the particular aspect of the
issue that a researcher is drawn to investigate. In this study, a holistic approach was used
in which the data from each participant were kept intact rather than identifying
commonalities or themes across the raw data from participants. An advantage of
preserving the diachronic nature of the events occurring in the period of time between
symptom onset and hospital arrival is that the connections between the parts of the story
and the relationship of the parts to the outcome are made explicit (Polkinghorne, 1995).
However, few detailed procedures were found in the methodological literature for a
holistic analysis of narrative data.
The aim of this paper is to describe the procedures used to conduct a holistic analysis of
women’s stories about the onset of ischemic stroke. I begin the paper by providing a brief
overview of the narrative methodology chosen to guide this study. This is followed by a
framework developed by Lieblich et al. (1998) that conceptualises narrative analysis as
occurring along several dimensions, including the holistic dimension. The procedures used
in the holistic analysis of narrative data for the study are then described and issues of
trustworthiness in narrative research are addressed. I conclude with remarks about the
utility and limitations of a holistic approach to narrative analysis.

Explanatory narrative research

Polkinghorne (1988) saw stories as a form of pre-linguistic mental representation in which a
series of temporally linked events are unified into an integrated psychological phenomenon.
Constructing a story is a way that human beings organise perception, thought, memory and
action to makes events in human life understandable and meaningful (Robinson and Hawpe,
1986). Explanatory narrative inquiry is a type of narrative methodology in which the aim is
to arrive at a plausible explanation for why something occurred in the manner that it did
(Polkinghorne, 1995). This methodology places emphasis on the motivations of the
individuals in the story by exhibiting the connections between perceptions, motivations
and actions in order to arrive at a narrative explanation for a particular outcome
(Polkinghorne, 1995).

Downloaded from jrn.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on July 11, 2015
694 Journal of Research in Nursing 18(8)

Conceptual framework for narrative analysis

The methodological diversity that characterises narrative inquiry is reflected in a variety of
approaches to the analysis of narrative data. A framework developed by Lieblich et al.
(1998) categorises narrative analysis along the dimensions of ‘holistic versus categorical’
and ‘content versus form’ (Figure 1). In holistic approaches, an individual’s story is viewed
as a whole and the parts within it are interpreted in relation to other parts of the story
(Lieblich et al., 1998). With categorical approaches, themes or categories are identified across
the data from multiple research participants (Lieblich et al., 1998). On the content/form
dimension of the framework, researchers may focus on what happened in the story, why it
occurred and the role of other people (content) or on how the story is structured (form)
(Lieblich et al., 1998).
The Lieblich et al. (1998) schema offers four possible approaches to the analysis of
narrative data: holistic-content; holistic-form; categorical-content and categorical-form. A
holistic-content analysis of psychiatric nurses’ stories by Kelly and Howe (2007)
chronologically configured each participant’s story to show the influence of Gestalt
training on their development as a nurse and their current practice. An analysis on the
holistic-form dimension results in the identification of overarching story types as was
exemplified by a study of breast cancer symptoms in Punjab women that resulted in four
story types: ‘nothing serious’, ‘worry’, ‘suspicion’ and ‘gaining strength’ (Bottorff et al.,
2007). Reissman (1993) describes how she segregated the text of the ‘divorce talk’ of
women and men into ‘stanzas’, or a series of lines about a common topic, in a
categorical-form analysis in order to ascertain how the stories told by women and men
differed. Categorical-content analysis is familiar to many researchers as qualitative content
analysis, the aim of which is to seek commonalities in the experiences of participants (Morse
and Field, 1995).
Lieblich et al. (1998) observed that a narrative analysis need not fit neatly into one of the
four dimensions of their schema, and Holloway and Freshwater (2007) advocated
the creative use of analytic techniques in order to understand narratives from
different perspectives. To that end, combining analytic approaches is common in

Figure 1. Classification of narrative analysis. Adapted from Lieblich et al. (1998).

Downloaded from jrn.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on July 11, 2015
Beal 695

narrative research. For example, Robichaux and Clark (2006) combined a structural analysis
of critical care nurses’ stories of end of life issues (categorical-form) with the identification of
story types (holistic-form). Ayres (2000) identified four distinct story types (holistic-form) in
the narratives of caregivers in addition to extrapolating themes from the data (categorical-
content). Other researchers combined holistic-content and holistic-categorical methods to
showcase the interaction between the insights drawn from the holistic analysis and the
themes in the categorical analysis (Dole, 2001; Mcilfatrick et al., 2006).

Methods
The methods for the study are described in detail elsewhere (Beal, Stuifbergen and Volker,
2012). After Institutional Review Board approval was obtained, 10 community dwelling
women with a verified ischemic stroke diagnosis within one year were recruited through
fliers distributed by the staff at outpatient rehabilitation facilities and through newspaper
advertisement. There are no well-defined criteria for determining sample size in narrative
inquiry and sample sizes generally are small (Creswell et al., 2007). The sample size for this
study was set based on other studies using a holistic-content approach to narrative analysis.
One of the 10 women who agreed to participate developed medical problems and withdrew
from the study before data collection. The age range of the remaining nine women was 24–84
years. Four women were Caucasian, three were Hispanic, one was African American and
one described herself as of Native American and Caucasian ancestry.
Data were obtained through individual interviews during which stories of stroke onset
were elicited. The women were invited to tell the story of their stroke from the moment they
first noticed symptoms until they arrived at the emergency department. The interviews were
audio recorded and the recordings were transcribed verbatim. In addition to the audio files
and transcriptions, the data consisted of field notes written immediately after each interview
and an analytic journal kept during data analysis as an aid to critical reflection and decision
making.

Holistic-content analysis
The aim of the holistic-content analysis used in this study was to create a written account of
each woman’s story. A challenge in writing narrative accounts in that people often do not tell
stories in a linear manner (Lincoln and Guba, 1985) and a narrative researcher must ‘re-story’
the story told to them by the participant (Ollerenshaw and Creswell, 2000 as cited in Creswell
et al., 2007). This activity involves temporally ordering the events and actions that occurred in
the period of time under study and then interpreting how these story elements relate to one
another and contribute to the ending of the story (Polkinghorne, 2007: 483).
The five-step process of constructing the narrative accounts was adapted from
Polkinghorne (1995) and was iterative between the transcription, plot outline (described
below) and the developing narrative account. The first step was to identify the plot of
each woman’s story. A plot consists of temporally linked events and actions that an
individual considers significant to their story, and it transforms events and action into a
whole ‘by highlighting and recognising the contribution that certain events make to the
development and outcome of the story’ (Polkinghorne, 1988: 18–19). Each transcript was
coded to designate events and actions using Bal’s (1985: 13) description of event as ‘the
transition from one state to another state’ and the dictionary definition of action as ‘the

Downloaded from jrn.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on July 11, 2015
696 Journal of Research in Nursing 18(8)

process or condition of acting or doing or the exertion of energy or influence’ (http//

www.dictionary.oed.com).
The transcript then was read to identify and code supporting data elements. Although
Polkinghorne (1995) used the term ‘data elements’ to refer to any part of the story,
‘supporting data elements’ were operationalised in the current study as sentences and/or
phrases that illuminated the context in which events and actions took place. For example,
consistent with Barton’s (2004) observation that illness is ‘nested’ within a life story, the
supporting data elements in the transcripts often revealed socially derived notions about
stroke and past experiences with personal and family illness that informed participants’
response to symptoms.
Although a holistic-content analysis primarily attends to the ‘who’, ‘what’ and ‘where’ of
the story (content), I also was drawn to the way that the stories were told (form). This was
particularly the case when the participants linked their experiences at stroke onset to another
episode in their lives. According to Polkinghorne (1988), associating a current experience
with a past experience is one of the cognitive processes involved in the construction of a story
that aides individuals to understand events in terms of a larger whole. An example of a
narrative linkage occurred when a participant was describing her emotional responses to her
symptoms and she began relating how frightened she had felt years earlier when one of her
children was injured. ‘Lisa’ eventually concluded that her stroke was the first time in her life
that she felt frightened for her own physical well-being. These and other narrative linkages in
the stories assisted the women – and me – to understand the events they were describing
within the context of their entire lives (Gubrium and Holstein, 1997).
The penultimate step of the within-case analysis was to construct a plot outline, which is a
visual representation of the participant’s story. On a large piece of paper, actions and events
were indicated in the order in which they occurred from left to right above a horizontal line
spanning the width of the paper. Supporting data elements and narrative linkages were
written below the line, corresponding to the relevant action/event. Plot outlines were
created on paper for the first five participants, but the amount of information was
extensive and proved cumbersome to transfer to paper form, and the remaining plot
outlines were computer generated. Box 1 depicts a portion of the plot outline developed
from ‘Maria’s’ story.
After the plot outline was completed, a narrative account of each participant’s story was
written. Each account began with observations about the first time I met each participant.
Then, I set down the plot of the story. Labov (1972) described the plot as the ‘skeleton’ of a
narrative, and I transformed the actions and events noted on the plot outline into sentences
in order to create a temporally ordered description of what occurred in the period of time
from when a woman first noticed her symptoms until she arrived at the hospital. This
ordering mirrored the basic story structure of beginning, middle and end (Polkinghorne,
1995). Then, the inclusion of supporting data elements and narrative linkages provided the
context for the actions and events that made up the plot, moving the analysis beyond an
ordering of what occurred in the period of time under study to a commentary that
‘uncovered and clarified the meaning of the story’ (Polkinghorne, 2007: 483). Box 2
contains an abridged version of ‘Maria’s’ narrative account that should allow readers to
follow the development of the plot.
The narrative accounts were written in my voice, reflecting that I had become the narrator
of the women’s stories (Polkinghorne, 1995). This placed me in the hermeneutic circle of
interpretation as I worked with each participant’s interpretation of her experiences to

Downloaded from jrn.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on July 11, 2015
Beal 697

Box 1. Experts from ‘Maria’s’ plotline. Note: E: Event; A: Action; I/P: Interpersonal Interaction.
Numbers after quotes correspond with the interview number and transcript line number.

E: Sudden, brief episode of arm weakness while in car.

‘I noticed it about 8:30 a.m. when I was in the car that my hand automatically dropped.’ 1 (4)
‘It was resting on the car door . . . but it just dropped.’ 1 (470)
‘When it dropped, it hit the corner of my door. And that’s when I got my first symptom.
That was the first thing of my stroke. Isn’t that weird?’ 1 (480)
‘I had it here [demonstrating hand laying on door edge] and it just dropped. I mean dropped.
There was no way of me to hold it up or anything.’ 1 (467)
‘And when it dropped it hit my funny bone.’
‘I thought it slipped while I was trying to go to sleep, and it hit my funny bone. That’s what
I related it to. No, I should not have done it to where, because I knew that what was really strange
for me, that it dropped. I mean it just . . . like a sack.’ 1 (471–476)
‘I didn’t think nothing. I was kind of curious but I just thought, no, it’s just, I was asleep and my
hand dropped.’ 1 (17–19)
CONTEXT: Maria was en route from her home to a city 60 miles away to a dance competition at the
time this occurred. Her husband was driving. They own a dance studio and often go to competitions
together.
A: Rubbed elbow because when arm dropped her elbow hit corner of door
‘I, you know, rubbed my elbow.’ 1 (6)
I/P: Told husband about hand dropping
‘And when it [funny bone feeling] went away I says, ‘‘You know, honey, it just dropped’’.’
‘And I told my husband it was weird.’
‘He said, ‘‘Are you sure you’re OK?’’ and I said, ‘‘Well, yea, look. I’m fine’’. So we went on.’ 1 (7–9)
‘He [husband] told me, ‘‘Do I need to go back. Do we need to turn back?’’ And I said ‘‘No,
everything is OK’’.’ 1(487)

produce my own interpretation of the story (Sandelowski, 1991: 165). Both the original
stories told by my participants and my re-storied narratives were constructions. My intent
was not to arrive at an objective account of what occurred but to locate each woman’s
experiences in time and place and focus on the meaning and consequences of their actions
(Bruner, 1986).
I entitled the accounts with a pseudonym chosen by the participant. A representational
quote from the interview text that epitomised each woman’s experiences was placed beneath
the pseudonym. ‘I can make it’, the quote I chose for ‘Maria’, reflected her determination to
control what was happening to her at stroke onset as she instructed her husband to drive
60 miles to their hometown hospital instead of seeking care at a nearby medical centre.

Comparison of narrative accounts

A collection of narrative accounts is an opportunity to apprehend the ‘the differences and
diversity of individuals and their storied experiences’ (Kelly and Howe, 2007: 141). However,
a group of stories, though rich in its portrayal of individual experience, is limited in its ability
to illuminate a more general understanding of the topic under study. Therefore, I conducted
a comparison of the narrative accounts in order to identify similar and dissimilar
characteristics of participants’ experiences at stroke onset (Polkinghorne, 1995). Rather
than extract themes from the raw data as in a categorical-content analysis, similarities
were identified across the narrative accounts because configuring each participant’s story

Downloaded from jrn.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on July 11, 2015
698 Journal of Research in Nursing 18(8)

Box 2. Excerpt from Maria’s narrative account.

It seemed as though I was barely in the door of the dance studio owned by Maria and her husband,
Craig, when Maria started to tell me the story of her stroke. Despite right sided weakness, Maria often
would rise from her chair during our interview to demonstrate how her body had acted on the day of
her stroke. Her gestures, and the fact she spoke rapidly with emphasis and animation, made it seem as
though this 55 year old Hispanic woman was enacting her story rather than telling it.
Maria and her husband had just set out for a dance competition when Maria’s right arm suddenly
dropped from where it was propped against the car door, causing her elbow to hit the door handle and
jolting her with an intense ‘funny bone’ sensation. At first, Maria wondered if she dozed off and her arm
had slipped, but after the ‘funny bone’ feeling passed she started thinking . . . Maria knew she was at risk
for stroke, and she had cared for both her parents when they had strokes. She wondered if this
symptom indicated a stroke. ‘I always had that in the back of my mind’, she said . . .
Craig asked if she wanted to turn back and be checked out by a doctor, but Maria said no. She was
reassured that her arms were symmetrical when she held them out, and her right arm felt as strong as
her left. She continued to test her arm periodically during the 60 mile drive to the competition.
When the couple arrived at the competition, Maria told Craig that she thought all was well because
she was walking and talking normally, and her arm appeared fine . . . The rest of the morning passed
uneventfully until around noon, when Maria developed a ‘terrible headache that just came on’. She took
two aspirin and closed her eyes and relaxed in her chair.
About a half hour later, Maria stood up to go to the restroom and realized she was unable to stand
up straight. ‘I was to the right . . . When I would try to straighten myself up, my body still kept on going
that way. It just tilted. It did not want to get straight’, she said. Maria described the sensation of leaning
to one side as ‘odd’ and ‘weird . . .’ and she sat back down to think. After a few minutes she reached the
conclusion that she was having a stroke because her mother had had the same symptom with a stroke.
As she had done that morning in the car, Maria decided to assess what was happening with her
body . . . ‘When I was walking I was, you know, kind of limping . . . I felt like I was short on one foot’, she
said, demonstrating to me how she was ‘unbalanced’ when she tried to walk . . . After taking a few
steps, Maria decided it wasn’t safe to walk and she sat back down and asked a student to get her
husband. They quickly decided she had to go the hospital . . . When she got into the car, Maria decided
to take two more aspirin ‘because I knew that I had a stroke’. She believed that aspirin would ‘stop a lot
of the damage’. A few minutes later, a disturbing thought occurred to her about the aspirin she had just
taken. ‘Then I remembered that too much aspirin could cause bleeding because that’s a blood thinner’,
she said. ‘But I thought that’s okay. I took it. I can’t do nothing about it. So . . . I just calmed myself . . . I
just kept myself calm because I thought if it’s my blood pressure, I don’t need my blood pressure going
up. You see, because blood pressure causes strokes, too’, she said.
. . . As they were pulling out of the parking lot Craig asked Maria to which of the two nearby
hospitals he should drive. Maria replied that she wanted the security and familiarity of her
hometown hospital . . . Her husband’s welfare figured into Maria’s choice to bypass local hospitals. ‘I
knew they were going to leave me at the hospital . . . and I was going to be there for weeks . . . If I had to
go in the hospital, it is nonsense him driving 60 miles every day or staying with me every day over
there . . .’
Craig immediately expressed concern about the wisdom of this plan . . . Maria argued that it was safe
to take the time to drive an hour to the hospital because she was still talking and thinking clearly. She
reasoned that if her thought processes were not affected then she was not in immediate danger . . . The
idea that stroke could be associated with not thinking clearly, and that this would be a sign that
immediate medical attention was necessary, came from Maria’s experience with her mother and her
sister. ‘When my mother had her stroke, and my sister, they couldn’t think clearly. You could see in
their eyes’. Reflecting on the difference between her symptoms and those of her mother and sister,
Maria concluded, ‘What else could happen? That’s how I looked at it . . .’.

Downloaded from jrn.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on July 11, 2015
Beal 699

into a coherent whole was central to an understanding of the phenomenon under study (Pak,
2006). Without this prior organisation, the links between various events and the overall
development of the events in the period of time under study would not be as apparent
(D. Polkinghorne, 28 January 2010, personal communication).
The procedure to compare the narrative accounts was based on qualitative content
analysis (Morse and Field, 1995). First, the narrative accounts were read and re-read in
order to obtain an overall impression of the women’s experiences during stroke onset.
Because of the complexity of the accounts, which in some cases described events
occurring over several days, a structure to organise the comparison was needed. Dodd
et al. (2001) conceptualised symptom experience as consisting of an individual’s
perception of a symptom, evaluation of the cause of the symptom and responses to the
symptom, and different coloured highlighters were used to identify text corresponding to
each of the three components of symptom experience. A fourth colour was used to identify
the actions and contributions of other people in the women’s stories because the role of other
people spanned all three components of symptom experience.
The next step in the comparative process consisted of identifying categories within each of
the (now) four components of symptoms experience. This involved further reading of the
narrative accounts, making notes in the margins to capture initial impressions. From these
notes, key concepts were identified. As the process continued, key concepts were organised
under larger labels. Eventually, between three and five labels were identified for each of the
four components of symptom experience, which became the categories that captured the
essential attributes of participants’ experiences at stroke onset.
After the categories were established, the relationship between the categories was analysed
across the four components of symptom experience. This was an important part of the
process because experiences in one component of symptom experience often were related
to experiences in another component. For example, a participant’s first interpretations of a
symptom as an everyday bodily sensation served as the basis for her actions in response to
symptoms, but as time went on and symptoms persisted or worsened or new symptoms
appeared, these bodily changes were re-interpreted and new actions taken based on the
revised interpretations. This step also entailed identifying differences in the accounts and
the contextual factors that may have accounted for the differences.
The comparison of the narrative accounts culminated in a written analysis and synthesis
of the findings. This document was organised under the headings of the four components of
symptom experience that had served as the organising framework for the comparison of the
accounts. The categories were thoroughly discussed, similarities and differences in the
narrative accounts as they related to the categories were described and the contextual
factors that contributed to these similarities and differences were explicated. Previous
research was brought forth in order to illustrate how the findings either were supported
by or diverged from this literature.

Trustworthiness of findings
When discussing the concept of validity in narrative research, Mischler (1990: 420)
distinguished the concept of ‘truth’, which assumes an objective reality, and
‘trustworthiness’ in which scientific knowledge is constructed through praxis whereby
readers decide whether they would use the findings as the basis for their own work.
Grounded in the work of Lincoln and Guba (1985), Graneheim and Lundman (2004)

Downloaded from jrn.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on July 11, 2015
700 Journal of Research in Nursing 18(8)

described credibility, dependability and transferability as aspects of trustworthiness that

researchers may develop to persuade readers that a qualitative research report is an
accurate reflection of what occurred in the study. For a narrative study, readers must ask
themselves whether the interpretations reached by the researcher adequately explain how the
events under study unfolded and whether a convincing explanation for the outcome of the
story had been achieved (Polkinghorne, 2007).
When appraising the credibility of a qualitative research report, a reader must decide
whether the methodology is appropriate given the problem statement and study purpose
(Morse and Field, 1995). The rationale for choosing explanatory narrative inquiry to guide
the current study was based in part on the tenet that story is a primary way that people
construct meaning in their lives (Bruner, 1986). The philosophical underpinnings of narrative
inquiry were described in as much detail in the original publication as could be accomplished
given space limitations. The methods used to choose the sample and collect and analyse the
data were fully described so that readers could assess the extent to which the findings were
based on data relevant to the study purpose (Graneheim and Lundman, 2004; Mirriam,
2009).
Credibility also concerns decisions made during data analysis and how these are presented
in the report (Graneheim and Lundman, 2004). To provide support for my interpretations of
the data, included in each narrative account were ‘particular description’ (quotes from the
transcripts), ‘general description’ (patterns discovered in each participant’s story) and
‘interpretive commentary’ (higher level of abstraction as the researcher reflects upon the
meaning of the story) (Pak, 2006). Quotations were used so that each woman’s own
words would be an integral part of the re-storied narratives. In order to articulate the
thought processes linking the data to my interpretations, I included first-person reflections
in the narrative accounts to open my thought processes to the scrutiny of readers.
Meetings with an experienced qualitative researcher during data collection and analysis
were a way to reduce potential bias, contributing to the trustworthiness of the research
report (Kahn, 2000). The aim of these meetings was not to obtain consensus on the
interpretation of the data, which would violate an underling assumption of qualitative
research that individuals may see reality differently from one another (Sandelowski, 1993).
The consulting researcher examined the transcriptions from two research participants and
the corresponding plot outline and narrative accounts in order to evaluate the extent to
which I hewed to the methodology and arrived at a plausible account of the series of events
leading up to hospital admission. This is consistent with Polkinghorne’s (2007) assertion that
a narrative researcher must convince readers what she or he is claiming about the meaning of
life events for the participants is reasonable given the evidence.
In order for a researcher’s interpretations to be dependable, she must be open to evolving
insights about the topic under study (Clandinin and Connelly, 2000; Lincoln and Guba, 1985).
Journaling allowed me to reflect upon the research process and document analytic decisions
(Rodgers and Cowles, 1993). The journal also served as a place to consider alternate
explanations for participants’ actions. Because it is inevitable that data analysis is shaped
by the researcher’s previous knowledge on the topic under study and their life experiences and
values (Lincoln and Guba, 1985; Polkinghorne, 2007), presenting alternatives in the narrative
accounts countered any tendency to resolve ambiguities in the data by choosing one
explanation over another when the meaning of a participant’s action was unclear.
The relevance of qualitative research for praxis is an aspect of validity in that nurses
must have confidence in the accuracy of the findings if they are to be used in practice

Downloaded from jrn.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on July 11, 2015
Beal 701

(Porter, 2007). In addition, deciding whether the findings accurately describe the experience
under study, nurses also must assess the applicability to their patients. One way to provide
evidence for transferability is to include ‘rich details and revealing descriptions’ in the findings
(Polkinghorne, 2007). Included in the research report were descriptions of the activities that
the women were engaged in at the time of the stroke as well as personal and family experiences
with illness. These elements, along with a description of the sample and the setting, placed the
events leading to hospital admission in context, contributing to the reader’s ability to assess
the utility of the study for practice (Graneheim and Lundman, 2004).

Discussion
A holistic-content approach to data analysis is congruent with the theoretical and
philosophical underpinnings of narrative inquiry. Bruner (1986) conceptualised narrative
thinking as a specific form of cognition in which people are naturally disposed to search
for the connections between events and actions. A holistic-content analysis, with its central
activity of configuring events and actions into a meaningful whole, is consistent with the
nature of the cognitive processes that people use to make meaning of life experiences as they
construct stories about their experiences. It is also consistent with an underlying assumption
of narrative inquiry that human experience occurs within a personal, social and cultural
sphere of understanding (Polkinghorne, 1988). The findings from this investigation revealed
that individual variations in participants’ life experiences were salient for the disparity in the
length of time (1 h to 6 days) that elapsed between first notice of symptoms and arrival at the
emergency department.
Temporality is at the heart of narrative inquiry. Ricoeur (1979) saw narrative as the ‘mode
of discourse through which the mode of being which we call temporality . . . is brought into
language’ (p. 17). Because time shapes events, narrators must ‘reckon with time’ when telling
a story and it is through narration that events in human lives become meaningful (Ricoeur,
1979). As do the participants in a narrative inquiry, narrative researchers reckon with time in
the process of chronologically ordering events and actions in the course of fashioning the
stories into narrative accounts.
Narratives do more than place actions and events in time. Telling a story involves
Heidegger’s (1962) level of temporality called ‘historicality’ in which time is experienced as
a ‘back and forth’ between the past, the ‘everyday-present-at-hand’ and what is yet to be
(Ricoeur, 1979). Every women in this study linked what happened to her at stroke onset with
past events, such as a previous personal illness or that of a family member, and when ‘re-
storying’ their stories I was able display how past events related to symptom response.
Holistic analysis may be intrinsically familiar and of value to nurses because it has
similarities with the ways nurses communicate with one another about clients in the
course of professional interactions (Simons and Squire, 2008). In addition, nurses know
that questions to patients and research participants frequently elicit stories. Based on
these stories, nurses reconstruct health histories and gain understanding of the meaning of
illness for patients by employing, albeit unconsciously, narrative modes of thought. A
holistic-content analysis is consistent with the relational aspect of storytelling wherein a
story is told with a listener rather than to a listener, thereby creating or shaping a
relationship with another individual (Frank, 2000: 354). Holistic-content analysis is an
extension of the nurse–patient relationship as the researcher extends the activity of
listening to a story into the analytic process (Chase, 2008: 73).

Downloaded from jrn.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on July 11, 2015
702 Journal of Research in Nursing 18(8)

Despite the utility of holistic-content analysis, this approach has limitations when used as
the sole method of analysis. Although it allows researchers to capture the intricacies of
individual experience, a holistic approach precludes elucidation of commonalities in the
data from multiple research participants (Ayres et al., 2003). Following the holistic
analysis with a comparison of the narrative accounts preserves the temporal dimension of
participant’s stories and the particulars of individual experience yet also enables researchers
to draw naturalistic generalisations (Sandelowski, 1996). The two-step approach to narrative
data described in this paper illuminated how the same general factors (e.g. the onset of
neurological symptoms) within the context of varying factors (e.g. differences in
socioeconomic status and life situation; prior experiences with illness; familiarity
with everyday bodily occurrences; the role of other people in health-decision making)
led to variations in the same general outcome (e.g. the timing of hospital arrival)
(Sandelowski, 1996).

Conclusion
The strength of a holistic-content approach to narrative data is that it takes into
consideration individual variations in an experience common to the participants. A
limitation of holistic-content analysis is that it does not result in a general understanding
of the topic under study. A structured comparison of the narrative accounts that result from
a holistic-content analysis is needed to identify similar and dissimilar characteristics of
participants’ experiences. This two-step approach to narrative data would be particularly
useful when a temporal ordering of events is relevant to understanding a problem or issue. In
addition to the exploration of acute and chronic illness trajectories, topics that may be well
suited to a holistic-content approach include health care decision making and nursing
professional development.

Key points for policy, practice and/or research

. Researchers using a holistic-content approach to narrative analysis create an account
that exhibits the connections between the parts of an individual’s story that provides
an explanation for the story outcome.
. The strength of a holistic-content approach is that it illuminates individual variations
in an experience common to research participants but an important limitation is that it
does not result in a general understanding of the topic under study.
. A structured comparison of the narrative accounts resulting from a holistic-content
analysis is needed to identify similar and dissimilar characteristics of participants’
experiences.
. A holistic-content approach to narrative data is especially useful when exploring time-
bounded phenomena including the onset of an acute illness.

Declaration of conflicting interest

None declared.

Downloaded from jrn.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on July 11, 2015
Beal 703

Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-
profit sectors.

References
Author (2012) A narrative study of women’s early symptom Gubrium JF and Holstein JA (1977) The New Language of
experience of ischemic stroke. Journal of Cardiovascular Qualitative Method. New York, NY: Oxford University
Nursing 27(3): 240–252. Press.
Ayres L (2000) Narratives of family caregiving: Four story Heidegger M (1962) Being and Time (Macquarrie J and
types. Research in Nursing & Health 23: 359–371. Robinson E, trans). New York, NY: Harper & Row.
Ayres L, Kavanaugh K and Knafl K (2003) Within-case and (originally published in 1927).
across-case approaches to qualitative data analysis. Holloway I and Freshwater D (2007) Narrative Research in
Qualitative Health Research 13: 871–883. Nursing. Oxford, UK: Blackwell Publishing.
Bal M (1985) Narratology: Introduction to the Theory of Kahn DL (2000) Reducing Bias. In: Kahn DL, Cohen M and
Narrative. Toronto, Canada: University of Toronto Press. Steeves RH (eds) Hermeneutic Phenomenological Research.
Barton SS (2004) Narrative inquiry: Locating aboriginal Thousand Oaks, CA: SAGE Publications, Inc, pp. 85–92.
epistemology in a relational methodology. Journal of Kelly T and Howe L (2007) Working with stories in nursing
Advanced Nursing 45(5): 519–526. research: Procedures used in narrative analysis. International
Beal C, Stuifbergen A and Volker D (2012) A narrative study Journal of Mental Health Nursing 16: 136–144.
of women’s early symptom experience of ischemic stroke. Labov W (1972) Language in the Inner City; Studies in the
Journal of Cardiovascular Nursing 27(3): 240–252. Black English Vernacular. Philadelphia, PA: University of
Bottorff JL, Grewal SK, Balneaves LG, et al. (2007) Punjabi Pennsylvania Press.
women’s stories of breast cancer symptoms: Gulti (lumps), Lieblich A, Tuval-Mashiach R and Zilber T (1998) Narrative
bumps, and darad (pain). Cancer Nursing 30(4): E36–E45. Research: Reading, Analysis and Interpretation. Thousand
Bruner J (1986) Actual Minds, Possible Worlds. Cambridge, Oaks, CA: SAGE Publications, Inc.
MA: Harvard University Press. Lincoln YS and Guba EG (1985) Naturalistic Inquiry.
Bruner J (1990) Acts of Meaning. Cambridge, MA: Harvard Newbury Park, CA: SAGE Publications, Inc.
University Press. Mcilfatrick S, Sullivan S and McKenna H (2006) Nursing the
Chang K, Tseng M and Tan T (2004) Prehospital delay after clinic vs. nursing the patient: Nurses’ experience of a day
hospital chemotherapy service. Journal of Clinical Nursing
acute stroke in Kaohsiung, Taiwan. Stroke 35: 700–704.
15: 1170–1178.
Chase SE (2008) Narrative inquiry. In: Denzin N and Lincoln
Memis S, Tugrul E, Didem E, et al. (2008) Multiple causes for
YS (eds) Collecting and Interpreting Qualitative Materials.
delay in arrival at hospital in acute stroke patients in Aydin,
Thousand Oaks, CA: SAGE Publications, pp. 57–91.
Turkey. BMC Neurology 815: 1–6.
Centers for Disease Control and Prevention (2008) Awareness
Mirriam SB (2009) Qualitative Research: A Guide to Design and
of stroke warning symptoms – 13 states and the District of
Implementation. San Francisco, CA: Jossey-Bass.
Columbia, 2005. MMWR 57(18): 481–485.
Mischler EG (1990) Validation in inquiry-guided research: The
Clandinin DJ and Connelly FM (2000) Narrative Inquiry:
role of exemplars in narrative studies. Harvard Educational
Experience and Story in Qualitative Research. San Francisco,
Review 60(4): 415–442.
CA: Jossey-Bass Publishers.
Morse J and Field P (1995) Qualitative Research Methods for
Creswell J, Clark V, Gutmann M, et al. (2003) Advanced
Health Professionals, (2nd ed). Thousand Oaks, CA: SAGE
mixed research designs. In: Tashakkori A and Teddlie C
Publications.
(eds) Handbook of Mixed Methods in Social & Behavioral
Research. Thousand Oaks, CA: SAGE Publications, Pak JH (2006) Korean American women: Stories of
pp. 209–240. acculturation and changing selves. Hoboken, NY: Routledge.
Dodd M, Janson S, Facione N, et al. (2001) Advancing the Polkinghorne DE (1988) Narrative Knowing and the Human
science of symptom management. Journal of Advanced Sciences. Albany, NY: State University of New York Press.
Nursing 33(5): 668–676. Polkinghorne DE (1995) Narrative configuration in qualitative
Deng YZ, Reeves MJ, Jacobs BS, et al., for the Paul Coverdell analysis. In: Wisniewski R and Amos HJ (eds) Life History
National Acute Stroke Registry Michigan Prototype and Narrative. London, UK: RoutledgeFalmer, pp. 5–24.
Investigators (2006) IV tissue plasminogen activator use in Polkinghorne DE (2007) Validity issues in narrative research.
acute stroke: experience from a statewide registry. Neurology Qualitative Inquiry 13: 471–486.
66(3): 306–312. Porter A (2007) Validity, trustworthiness and rigour:
Dole S (2001) Reconciling contradictions: Identity formation Reasserting realism in qualitative research. Journal of
in individuals with giftedness and learning disabilities. Advanced Nursing 60(1): 79–86.
Journal for the Education of the Gifted 25(2): 103–137. Reissman CK (1993) Narrative Analysis. Thousand Oaks, CA:
Frank AW (2000) The standpoint of storyteller. Qualitative SAGE Publications, Inc.
Health Research 10: 354–365. Ricoeur P (1979) The human experience of time and narrative.
Graneheim UH and Lundman B (2004) Qualitative content Research in Phenomenology 9: 17–34.
analysis in nursing research: Concepts procedures and Robichaux CM and Clark AP (2006) Practice of expert critical
meausres to achieve trustworthiness. Nurse Education Today care nurses in situations of prognostic conflict at the end of
24: 105–112. life. American Journal of Critical Care 15(5): 480–489.

Downloaded from jrn.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on July 11, 2015
704 Journal of Research in Nursing 18(8)

Robinson JA and Hawpe L (1986) Narrative thinking as a Sandelowski M (1996) One is the liveliest number: The case
heuristic process. In: Sarbin T (ed.) Narrative Psychology: orientation of qualitative research. Research in Nursing and
The Storied Nature of Human Conduct. New York, NY: Health 19: 525–529.
Praeger, pp. 111–125. Simons L and Squire C (2008) Shifting the focus: Sequential
Rodgers BL and Cowles KV (1993) The qualitative research methods of analysis with qualitative data. Qualitative Health
audit trail: A complex collection of documentation. Research Research 18(1): 120–132.
in Nursing and Health 16: 219–226. Smith MA, Lisabeth LD and Bonikowski F (2010) The role of
Sandelowski M (1991) Telling stories: Narrative approaches in ethnicity, sex, and language on delay to hospital arrival for
qualitative research. IMAGE: Journal of Nursing Scholarship acute ischemic stroke. Stroke: A Journal of Cerebral
23(3): 161–166. Circulation 41(5): 905–909.
Sandelowski M (1993) Rigor or rigor mortis: The problem of
rigor in qualitative research. Advances in Nursing Science
16(2): 1–8.

Claudia C. Beal (PhD, RN) is an assistant professor at Baylor University, Louise Herrington
School of Nursing in Dallas, Texas, USA where she teaches applied clinical ethics, research
methods and translation of evidence courses in the graduate program. Her professional
interests are broadly in the area of women’s health with a focus on chronic and disabling
conditions. Claudia has a particular interest in stroke and in several studies has examined the
symptom management of acute ischemic stroke. In her current study Claudia is examining
stroke health literacy among African American women. Claudia received a PhD from the
University of Texas at Austin where she completed a portfolio in women and gender studies
with a specialisation in women and health through the Center for Women and Gender
Studies. Claudia’s work has been published in leading nursing journals and she serves as a
manuscript reviewer for several professional journals.

Downloaded from jrn.sagepub.com at UNIV CALIFORNIA SANTA BARBARA on July 11, 2015