3. Methodology
3.3 Data Analysis
Data obtained in this study was more varied and complex than quantitative data since meanings and realities are dependent on the participant’s individual interpretations of situations that occur around them, subsequently generating data that are likely to be large in volume and complex in nature (Saunders et al., 2016). This required me to thoroughly analyze the richness of the data to explore subjects in as real a manner
RQ: How do adolescent’s construct their identity through the symbolic consumption of Smartphones?
SQ1 How do adolescent’s subjective experiences with Smartphones affect their identity construction?
SQ2 How do adolescents understand the symbolic meanings attached to Smartphones?
SQ3 How do adolescent’s express their identity through their individual choice of Smartphone?
Giddens (1991; 1996)
Erikson (1968) Noble & Walker (1997) Van Gennep (1960) John (1999)
Kleine et al. (1995)
• Tænker du over hvilken telefon du har?
• Tænker du over, hvad andre synes om dit valg af telefon?
• Bliver du påvirket af dine venner, når du skal have ny telefon?
• Hvordan ville du have det, hvis du var nødt til at låne en ældre telefon I et stykke tid?
• Hvordan vil du gerne have, at dine venner opfatter dig?
• Har dine ønsker for telefoner ændret sig efter du er blevet ældre?
• Er du blevet mere bevidst om hvad din telefon fortæller om dig?
• Er der kommet mere fokus på telefoner efter din mening?
• Føler du, at du passer bedre ind I din vennegruppe, når du har netop den her telefon?
• Gør det dig mere selvsikker at have den her telefon fremfor en ældre model?
• Hvordan skelner du mellem om en telefon passer til din stil?
• Hvad afgøre at en telefon er cool?
• Tænker du over, at din telefon skal passe til resten af din stil?
• Hvilken rolle spiller din telefon i dit liv?
• Kunne du undvære din telefon?
Elliott Wattanasuwan (1998)
Levy (1959)
Solomon (1983) McCracken (1986; 1988)
• Hvilke features på din telefon sætter du mest pris på?
• Kan du finde på at skifte din telefon selvom den ikke er gået i stykker?
• Hvad er det vigtigste for dig, når du skal have en ny telefon?
• Tænker du over hvad du udstråler med dit valg af telefon?
• Er det vigtigt for dig, hvordan din telefon ser ud / hvilket brand det er?
• Kan du forestiller dig, at det i dag er nogen uden en Smartphone?
• Er der status i at have en bestemt type telefon?
• Er det vigtigt for dig at have den samme telefon som dine venner?
• Føler du, at du er mere en del af fællesskabet fordi du har den her telefon?
• Tilhører man en bestemt gruppe på grund af sin telefon?
Sørensen & Thomsen (2006)
Richins (1994)
Belk (1988) Sartre (1943)
• Prøv at sæt nogle ord på, hvordan du bruger din telefon til at udstråle hvem du er?
• Er det vigtigt for dig, at have nogle ting der tydeligt afspejler dig?
• Fungere telefoner som et symbol til andre om hvilken type du er?
• Hvordan oplever du det at have en telefon?
• Hvad er det der gør forskellen på en Smartphone og en ikke-Smartphone?
• Hvordan ville du have det, hvis du ikke havde din telefon?
• Ville der mangle noget af dig, hvis du skulle undvære din telefon i længere tid?
• Er din telefon en del af dig?
• Hvad udstråler du til dine venner med den telefon du har noget sammenlignet med den du havde tidligere?
• Betyder de de ting du kan gøre med din telefon meget for dig?
Reflective Processes Ideal self Conscious vs.
unconscious behavior
Role Transition Strive for Belongingness Consumer Socialization
Attachments
“Me-ness”
Intangible Attributes Rational vs. irrational decision criteria Language of Symbols
Social Symbolism Social Comparisons Social Roles and Groups
Signal vs. Experiential Value
Private vs. Public Domain
Extended Self Loss of Possessions Enable Having, Doingand Being
Sub-question Theory Question Categories Intended Interview Questions
Table 3: Interview guide overview (Source: Own Construction)
as possible (Ibid.). As meanings were essentially derived from interpretations that could contain multiple meanings, analyzing data entailed exploring and clarifying these with great care, in which I considered the research approach and how this affected the subsequent analysis. Given the predominant deductive approach, I aimed to use existing theory to shape my analysis, where certain advantages helped me link the research into the existing body of literature, contributing with new and insightful knowledge (Ibid.). The process of data collection, data analysis as well as the development and verification of propositions was an interactive set of processes that, in turn, was undertaking both during and after data collection (Ibid.). This involved a process where I coded and categorized data in order to group them according to the themes that appeared once I began to make sense of the data. In turn, these themes provided structure to answer the research question. Nevertheless, before engaging in the process of coding, a brief assessment of how data was recorded and transcribed is outlined.
3.3.1 The Process of Recording and Transcribing Data
Working the interviews were often inhibit by the demands of conduction both semi-structured and focus group interviews, because of the potential flow of new ideas that often emerge during the interviews (Saunders et al., 2016). Thus, to optimize the value gained from the interviews, I chose to compile a full audio-record with permission from all participants to ensure a subsequent detailed understanding of what was expressed, as well as avoid mixing up data from the different interviews (Ibid.). This further allowed me to re-listen to the interviews during the data analysis, which enabled me to reinforce my understanding of the data and identify quotes to demonstrate the main points (Ibid.).
I decided not to work directly from the recording, but instead to produce a transcription of the recording.
This process of transcription was where the interviews were transformed from spoken language to written language (Saunders et al., 2016). This is typically a necessary step in qualitative research on the way to making valid interpretation and to ensure relevant data is not lost (Flick, 2014; Saunders et al., 2016). It required me to make decisions about the exactness of the transcriptions, to eventually formulating rules to achieve consistency and comparability among the interviews (Flick, 2014). The level of details perceived required for this research followed the social scientific qualitative transcription tradition advocated by Steinar Kvale, in which transcriptions are seen as interpretive constructions that are useful instruments for a given purpose rather than copies or representations of an original reality (Kvale, 1994). This is consistent with the logic proposed by Bruce (1992), who argued that “it is reasonable to think that a transcription system should be easy to write, easy to read, easy to learn and easy to teach” (Bruce, 1992 quoted in Flick, 2014). Because of this, elements of the interviews were omitted to invest more time and energy in interpreting data. However,
a significant part of the transcription process was capturing the meanings of utterances that were profoundly shaped by the way something was said in addition to what was said (Bailey, 2008). While transcription in this context needed to be more detailed to capture the features of talk such as emphasis, tone of voice, timing, and pauses, it was generally impossible to represent the full complexity of human interaction on a transcript (Bailey, 2008; Saunders et al., 2016). Thus, listening to the recordings throughout the data analysis brought data alive by appreciating the ways participants had expressed themselves as well as the way it was said (Bailey, 2008). Concomitantly it is noteworthy that the transcription was not just a meaningless technical and uncomplicated procedure; rather, it was a transformation of meaning from one medium to another, involving both analytical decisions and interpretive consciousness (Flick, 2014).
3.3.2 The Process of Coding
Reading and reflecting on the transcribed data required seeing across data and above the individual documents to identify themes and ideas. The goal of qualitative coding was to learn from the data as well as to keep revising data extracts until I realized and understood patterns in the data (Saldaña, 2015).
Considering the methodological propositions applied in this research, Thematic Analysis was used to focus on examining themes within data to identify implicit and explicit ideas within and across data with the aim of providing a rich thematic description, consequently giving the reader a sense of the predominant and important themes (Braun & Clarke, 2006). When identifying themes the process of coding needed to be an accurate reflection of the content of the entire data set. This means that some depth and complexity could be lost, yet a rich overall description was maintained (Ibid.). Add to this that thematic analysis offered a systematic yet flexible and accessible approach to analyzing qualitative data (Ibid.). It was systematic because it enabled an orderly way to analyze data, leading to rich descriptions, explanations, and theorizing (Saunders et al., 2015). Additionally, while its flexibility was seemingly conspicuous as it was not necessarily linked to or associated with any particular philosophical position, the aim of using it in interpretive research was to explore the different interpretations of the studied phenomenon (Ibid.). Themes and patterns were identified using a deductive, or theoretical, thematic analysis. This form of thematic analysis was not exclusively data-driven; instead, I tried to fit data into a pre-existing coding frame based on my theoretical interests in the area (Braun & Clarke, 2006). Thus, coding for a specific research question incorporated elements of an inductive approach that helped me to ensure a rigorous and rich description of the data set (Ibid.).
Another important decision revolves around the level of which themes are to be identified (Braun & Clarke, 2006). I sought to go beyond the semantic content of the data to identify and examine the underlying ideas
and assumptions that were theorized as shaping the semantic content of the data (Ibid.). Thus, the development of themes from a latent thematic analysis approach involved interpretive work in which I looked for something that went beyond what the participants explicitly expressed (Ibid.). Also, the analysis was not a linear process; rather, it was a recursive process, where I moved back and forth between phases to eventually produce a rich data analysis (Ibid.). As mentioned above, the first phase of my data analysis was concerned with transcription, in which I began to familiarize myself with the depth and breadth of the data containedin the process of immersion which continued throughout the analysis (Ibid.). Without being familiar with all aspects of my data collection, I would not have been able to engage in the analytical process appropriately, since it required me to re-read and evaluate data during the analysis (Saunders et al., 2016;
Braun & Clarke, 2006). Following this, I generated initial codes in a systematic manner across the entire data set, thereby collecting data relevant to each identified code (Braun & Clarke, 2006). This entailed working systematically through the data to provide complete and equal attention to each data item and identify interesting aspects in the data that could form the basis for repeated patterns (i.e. themes) across the data set (Ibid.).
My coding was performed through NVivo, a software program used to categorize and classify data. This made the process of coding more accessible and convenient because it enabled me to bring together and combine the collected data in a meaningful way, finding connections and relationships across the data set more easily (NVivo, n.d.). NVivo provided me with a platform to improve my data management by organizing, storing, and retrieving data throughout the entire data analysis, which enabled me to work more efficiently and rigorously back up findings with evidence (Ibid.). I began by collating relevant data for each code, subsequently sorting the different codes into potential themes. As I considered how the different codes may combine to form an overarching theme, the interpretive analysis of my data began (Boyatzis, 1998 cited in Brand & Clarke, 2006, p. 88). This was when I started to reflect upon the connection between codes and themes, thereby regarding some initial codes as main themes, whereas others would form sub-themes or be discarded (Braun & Clarke, 2006). This involved a refinement of themes during which it became clear which themes did not have enough data to support them, which ones might collapse into each other, as well as which themes needed to be broken down into separate themes (Ibid.). Based on this, I had a fairly good idea of which different themes were most applicable to the overall story as reported by respondents. The next step in the ongoing analysis was to generate clear definitions and names for each theme before producing the final analysis (see appendix 11 for identified themes and appertaining nodes). It was important to provide sufficient evidence of the themes within data by selecting enough compelling data extracts to demonstrate
the prevalence of the theme (Ibid.). Thus, the extracts applied in the final analysis should be easily identifiable as an example of a particular issue.