4.5 Main Study 126
4.5.6 Findings, Coding and Analyses 149
The analysis of the data started as early as the data collection stage. By doing so, it allowed me to understand the data recently gathered and to adjust later interview questions or prompts and the guidelines I had for the online observation. Only after the completion of the data collection stage (interview and observation) did the final data analysis commence. I gathered all the data from interviews and notes from the online observation to be re-‐read and re-‐analysed using the framework I set.
Data preparation and analyses were done using specific software – Express Scribe, NVivo, Snagit and also manually. Snagit was used to capture data online such as photos and pages. It allows for full window scrolling that captures data beyond those shown on the window. It was particularly useful in capturing Facebook pages and profile pages that have lengthy information that had to be scrolled down to be read50. This capability to scroll down Facebook profiles demonstrates the
affordances written about the site in a previous chapter. An online site such as
50 It is a proprietary software from TechSmith which is easily accessible and available for
Facebook, due to its data persistence, allows sharing previously made to be searched. These affordances the site offers were taken advantage of in this research. It allows earlier and current social interactions and sharing to be captured using a tool such as Snagit. Using these captured Facebook profiles and snippets of social interactions, observation and analyses of delayed and real-‐time interactions were conducted. Express Scribe was used to transcribe audio interviews, which were later integrated into NVivo for further coding and analysis. The use of software helped in organising the data into files and sections that are easily accessible and understood. There were however times when I felt too distant from my data, so that I resorted to manual analysis. Using coloured pens and paper helped to get me back to the data when the tedious work of doing the analysis on a computer removed the nearness and familiarity (Crang and Cook, 2007). By going back and forth between manual and software led analysis helped in clearing my thoughts and provided me with new ideas or angles to analyse the data.
I went through a number of coding stages using different strategies (Saldana, 2009). I started using the broad coding method to code the individual transcribed interviews. From a small number of free codes I managed to generate a large number of free codes that eventually made the coding too cumbersome. At this stage, I tried to get as many codes as possible that included: feelings, actions, attributes, influence and motives. From this I was able to generate a number of isolated code groups. Smaller nodes under each group were assessed individually and linked to nodes from other code groups to bring out their relationships. Breaking them apart and bringing them together helped to clarify the confusion and
the missing links between nodes and group codes. There were also times when I felt I had done the coding the wrong way and decided to start afresh. One of the advantages of NVivo is that all the codes generated earlier are saved even though new codes are created. This allowed me to go back and forth assessing the already created codes and the newly generated codes to see if there were any similarities or differences and if they should be put aside. There were a number of codes that were merged together because they represented the same thing. In this second coding stage, I moved my coding activity to the hard copy of interviews transcriptions, which I found to be very rewarding, especially after losing touch with the data on the computer. By removing myself from the computer and NVivo, transcribing the hard copy physically freshened my mind and helped to open up different angles to look at the data. Codes generated for the pilot interview also proved to be very valuable for coding the interviews from the main study. Although the earlier codes were not as comprehensive and detailed as those in the main study, having them close by while coding the later interviews made grouping codes and creating the names easier.
Coming back to the NVivo software, I conducted as much detailed coding as possible and went deeper to see variations in the earlier broad free codes I generated. Codes were moved around, deleted or created to better represent the data. Naming codes was a tedious task that required constant assessing and reassessing of the labels. Suitability of those codes attached to the data was one of my biggest concerns. Improper or irrelevant code names would affect my data analysis later, so that I had to think ahead of the coding stage to my data analysis
stage. To make sense of the data, I used the Model feature in NVivo that brought the codes together on one page. It allows the user to view all the codes created in the form of a mind map, with the flexibility of moving the codes around, changing the shapes, create relationships of code to another. With the model created, the codes and their relationships became clearer. I was able to see the emerging themes. The codes that were earlier in specific groups were moved around to more suitable groups. Other codes that were isolated, which might be valuable later, were left on the side.
This research is focused on online social interactions on Facebook, although it acknowledged online and offline interactions and experiences to be mutually constitutive, therefore the codes are arranged according to their online or offline activities. This allowed me to have a clearer picture of what is going on in offline and online spaces. Only by separating them in this way can I see the relationship between the two spaces.
Codes for online activities are grouped into a number of labels: Interactions Online; Managing Friendships; Family Relationship Online; Friends Relationship Online; Online Self; Positive Experience of Facebook; Negative Experience of Facebook; Openness due to Online Interactions; Tech Know How; Means of Communication; and Expectations on Facebook. Codes for offline activities are categorised into a number of groups: Descriptions of Malaysia; Background and Experience; Trust and Access; Inter-‐ethnic Relationships; Self-‐concept; Changing Personalities; Phase of Life; and Openness due to Overseas Travel.
These codes, even though they appear to be specific to offline or online environments, are inextricably linked to one another and often influence one another. For instance Phase of Life and Changing Personalities offline shape Online Interactions and Online Self. The relationships between them are mutually constitutive. They cover the respondents online and offline experiences and these experiences influence on the cultivation cosmopolitan sensibilities and performance of cosmopolitanism.
The first stage of analysis after coding focused on how Facebook is used in the respondent’s everyday life, the significance of the site, and their experiences with their own context. During the interview I received a lot of answers on how Facebook was used daily and from there it was apparent that the experiences and reflexivity process they went through while they were using Facebook (not necessarily everyday) as a continuum are complex, not always straight-‐forward but laden with dilemmas, thoughts, contradictions, inclusion and exclusion, ignorance and acceptance, and negotiation. They are to some extent confined within their own Malay Muslim contexts online, that resulted in their strategic performance of self. Their actual use of the site and what they do online are relevant to this cosmopolitanism study, to such an extent that they shape cosmopolitan consciousness and its performance. This is followed by the second stage of analysis that examined the coded data using the proposed framework outlined in chapter three: analytical tools for cosmopolitan sensibilities and cosmopolitan performance. From the different stages of data analyses, I have a number of interconnected
themes (see below) which are discussed throughout the empirical chapters of this thesis (Chapter 5 – 7) in relation to online social interactions and cosmopolitanism.
• staying true to self;
• Facebook as extension of self;
• negotiating everyday life away from home;
• social structures and socio-‐cultural and religious contexts brought online;
• practising and strengthening core values and beliefs online; • family relationships managed in different ways;
• the loose concept of friendship and interactions with others occurring differently than offline;
• strategically accepting others and strategic performance of open (not necessarily cosmopolitan) self.
The empirical chapters of this thesis are organised in this way: chapter five deals with the respondents’ experiences on Facebook, using the site while they are away from home (Malaysia). This chapter discusses the complexities of the respondents’ Facebook experiences by looking into their self-‐presentation and social relationships online, without any reference to cosmopolitanism at this stage. Their experiences of going online and being online set the background context of their online presence. Chapters Six and Seven address the core of this research; that is cosmopolitan sensibilities and cosmopolitan performance respectively. Chapter six engages with the re-‐thinking of cosmopolitanism using this term Rooted Muslim Cosmopolitanism labelled for their experience to explore and bring to the front
matters significant to (their) cosmopolitanism. This chapter draws out facets that are significant to this Rooted Muslim Cosmopolitanism, particularly the constant battle individuals experienced within their self, which eventually led to strategic deployment of religious discursive resources to navigate everyday online and offline life away from home. Chapter seven, building from the discussion left off in chapter six further deals with performance of those cosmopolitan sensibilities. It continues with the argument of cosmopolitanism as a strategy and demonstrated using empirical evidence how openness is performed as part of one’s strategy. It discusses a common view of any forms of rooted/ Islamic cosmopolitanism that is in marked contrast to liberal Western cosmopolitanism. Using performance of religiosity (hijab and intimacies on Facebook) this chapter argues that performance of religiosity is not the anti-‐thesis of any cosmopolitanism.
To end this section, there is also an important point, which I am compelled to highlight here. Their cosmopolitan experiences that I studied through interview and observation, are limited to a specific time-‐frame and the available settings and features during the time of field work. Interviews allowed users’ experiences to be elicited but what they shared are of the past based on what they remembered they did. Re-‐thinking about this, in the context of the fallibility of memory, what they shared might not be the actual cosmopolitanism felt and performed at that time and they could also be refraining from saying certain topics during the interviews (Busher and James, 2006). Based on my experience interviewing others and myself as an interviewee, what was said during interviews was not always straight-‐forward but shaped by the dynamic of the interview and the researcher/respondent
relationships as “all knowledge is produced in specific circumstances and that those circumstances shape it in some way” (Rose 1997: 305). Interviewees could refer to certain contexts at that time and provide examples suitable to express that when answering an interviewer’s questions. For example, when considering how one extends openness to cultural others, Kitzinger (2004: 128 cited in Silverman 2011: 181) has written “what (are said) should not be taken as evidence of their experience, but only as a form of talk – a ‘discourse’, ‘account’ or ‘repertoire’ – which represents a culturally available way of packaging experience”. At other times it could be a different (un-‐cosmopolitan) experience within the same context. Therefore contradictions are not always obvious and clear cut. They might only be palpable to the interviewee because those dilemmas, thoughts, and contradictions are in their mind. This, however, should not be seen only as limitations of this cosmopolitan research but as part of the complexities and challenges involved in doing fieldwork and social research. This research does not aim to provide a general statement summing up Malay Malaysian students’ (in the UK or in Malaysia) cosmopolitan experiences but to provide academia, particularly those relating to cosmopolitanism discourse, with new angles to study cosmopolitanism and to recognise that actual cosmopolitanism might be difficult to detect; in particular, that sometimes what is said by respondents could be a product of the past not the current experiences. Acknowledging this matter could provide this researcher (and scholars of cosmopolitanism studies) with some ideas to work with, especially in capturing real-‐time cosmopolitan performance when and where possible.