Chapter I: Introduction
Chapter 3: Research Methodology and Methods
3.4. Data Analysis
Participants’ narrative data collected in the present study was analyzed using a
combination of holistic content analysis (Lieblich et al.,1998) and thematic analysis (Braun and Clarke, 2006). The holistic content analysis involved viewing each participant’s story as a whole while focusing on the plot presented by the story during the analysis (Lieblich et al.,1998). The thematic analysis involved identifying common patterns and themes across four participants’
stories that are related to my research question. A primarily inductive (i.e., bottom-up) approach was utilized to analyze the narrative data. The data analysis process was also driven by my theoretical framework, the ecological system theory. The data analysis process took paces in four stages, which are described below.
3.4.1. Transcription and Data Cleaning. In preparation for the data analysis, each BNI session was transcribed immediately following its competition. The interviews were transcribed by a research assistant in order to accommodate the tight timeline of this master’s thesis project.
I acknowledge that by having a research assistant transcribe the interviews instead of
transcribing them myself, I may have missed the opportunity to be fully immersed in the data
and to discover emerging patterns and themes from participants’ stories prior to conducting the narrative analysis. Moreover, I am aware that participants’ changes in tone of voice and pauses in speech could not be fully captured within the interview transcripts. To compensate for this limitation, I listened to each audio recorded interview multiple times before performing data analysis (Lieblich et al. 1998). While listening to the audio recorded interviews, I also read through the corresponding transcripts to check their accuracy.
I then cleaned the interview transcripts. Specifically, major grammatical errors were corrected, and repeated words that did not express any meaning were removed, for the purpose of clarity. In order to protect the identity of those individuals mentioned within the transcripts, all names mentioned in the interview transcripts were replaced with pseudonyms. Among the two participants who took part in the feedback interview session, one chose the pseudonyms for herself and her child, and the other gave me permission to choose the pseudonyms for her and her child. For the two participants who I could not get in touch with after the second BNI session, I picked the pseudonyms for them and their children.
3.4.2. Holistic-Content Mode: Generating Themes from Individual Transcripts. This stage of the analysis began with familiarizing myself with the data, by listening to the audio recorded interviews and reading through the interview transcripts. During this active
reading/listening phase, I jotted down notes about the data (e.g., my interpretations of the significance of a particular life event to the participant, and my observations of the ways in which the participant made sense of an experience). In addition, I highlighted the content in the transcripts that I considered meaningful, interesting, or prominent in the participant’s story. The note-making process at this stage was causal and observational rather than systematic.
Once I became familiar with the data, I began to analyze the content systematically through generating initial codes. Codes are labels that highlight certain features of the data that are potentially relevant to the research question (Braun & Clarke,2006). An example of a code noted in many of the interview transcripts is promoting independence in adolescent child. The coding process was done by hand in a Microsoft Word document. The process involved identifying portions of the data that were significant and potentially relevant to my research questions, taking note of the content associated with each code through colour coding the text.
This process was repeated until the transcript was fully coded and all relevant content was marked. Some of the codes were descriptive, staying close to the story content and mirroring participants’ words; others were interpretive, integrating my interpretations and extending beyond what participants said.
Then I put the participant’s interview data from the two BNI sessions side by side, reviewing the coded data from both transcripts. Upon reviewing the data from two interviews as a whole, I combined the existing codes that shared similar unifying features into themes and identified the storylines of the narrative data. The within-story themes created were each assigned a title that conveyed their central meaning. All transcripts were coded following this procedure.
3.4.3. Re-writing the Narratives. The next step of data analysis involved re-writing the stories told by participants during the interviews into personal narratives. As noted by Lincoln and Guba (1985), people often do not tell their stories in a linear manner (Creswell, Hanson, Clark Plano, & Morales, 2007; Lieblich et al. 1998). Therefore, to start this narrative re-writing process, I re-examined each participant’s interview transcripts from the two BNI sessions and organized the story in chronological order. When re-writing the narratives, I focused on
participants’ contributions during the interviews, and used a combination of the themes that I had generated as well as the storylines that I had identified as a guide to frame the narratives. My theoretical framework, the ecological system theory, also came into play at this stage of analysis.
In particular, I highlighted how various components in participants’ lived environments (e.g., families, social service providers, the school system, cultural context) shaped their parenting experiences, as well as how they asserted influences on their environment as parents (e.g., advocating for their child). I took the role as a narrator and an observer in the re-constructed personal narratives and wrote those narratives from the third-person point of view.
3.4.4. Identifying Themes across the Data Set. After the central themes were generated from interview transcripts and participants’ personal narratives were re-constructed, I started to generate the overarching themes across data sets using thematic analysis (Braun & Clarke, 2006). Specifically, I explored the relationships between the themes that I had created and considered how this list of themes across stories could work together to tell “an overall story about the data” (Braun & Clarke, 2006, p. 65). I then examined the potential themes in relation to the coded data, the whole data set, and the research question. At this stage of the analysis, certain themes identified within individual participants’ narrative accounts were modified or discarded (Braun & Clarke, 2006). This final phase of data analysis generated six overarching themes, which will be presented in detail in Chapter 5.
My Master’s thesis supervisor and I both participated in this final stage of the data analysis process. After analyzing the narrative data from all participants’ interviews, I emailed the data analysis summary of each participant’s interview responses to my thesis supervisor, who had reviewed all the interview transcripts beforehand. Those summaries included the list of themes that I had generated from each participant’s story and corresponding exemplar quotes.
Next, my supervisor and I each read through all four summaries and generated the overarching themes across the data set. After we both completed this process on our own time, we held a two-hour meeting to discuss the major themes that we had identified. Towards the end of the meeting, we reached agreement regarding the most prominent themes emerging across the narrative data.
The purpose of having two researchers analyzing the data was not to a find the single ‘truth’ in the data. Instead, this practice, referred to as co-reflexivity, aimed to ensure that the data was analyzed rigorously and allowed the researchers to take a step back and reflect critically on the assumptions that they bring to the table (Moore, Noble-Carr, & McArthur, 2016).