The process of qualitative analysis is an ongoing and iterative one. Making meaning, and making connections, began as I was carrying out my interviews (as I had ideas and noted down impressions and emotions), influenced my data generation (as thoughts generated from previous interviews affected how I pursued particular lines of inquiry), and has continued through the process of writing (as particular ideas fall into place – or fail to – as I try to place them in a coherent order). As such, as
96 Amanda Coffey and Paul Atkinson suggest, analysis was not a “distinct stage ” of research (1996, p. 6), but was a reflexive activity embedded in the broader project. The period of interviewing spanned the school summer holidays, with some of the initial interviews carried out in the summer term, and some in the autumn. This gave me a significant period of time to think through the data I had already generated, and pursue connections in the forthcoming period of interviewing.
In categorising, coding, thinking through, dividing and connecting, I was trying to represent how young people were creating themselves as subjects in the specifically located space of the interview, the ways and extents to which they invested in particular gendered and classed discourses, as well as their representations of their own experiences. As Miller and Glassner (2004/1997) argue, it is not necessary to choose between form and content in the analysis of interviews; while always interpreting the interview as a situated form of interaction, I still analyse the interview talk for its mediated representation of experience, as I have discussed previously.
To aid in the process of formal analysis, I used qualitative analysis computer
software (CAQDAS). Specifically, after audio-recording and transcribing the text of all interviews, I entered the transcripts into NVivo 8. As many writers have
cautioned, CAQDAS is not capable of doing the thinking and is not a substitute for analysis, but it can be a helpful tool for aiding organisation, access and retrieval of data, and for developing themes and concepts through categorisation. I began by coding the data, looking for themes and patterns across the body of interviews. Throughout initial coding, I employed both “in vivo ” and “sociologically
constructed” codes (Strauss 1987), categorising both according to participants' own talk and use of language, and according to theoretical interpretations based on previous literature and related to my research questions. In this sense, analysing and interpreting participants' talk was also a process of going “beyond the data” (Coffey and Atkinson 1996; Silverman 2000), making linkages with wider thought.
97 The coding, categorisation, splitting and retrieval of data can create problems of decontextualisation, which may be exacerbated by the ease of these techniques in CAQDAS (Fielding and Lee 1998). A particular point made by a participant on a particular topic may be coded and retrieved alone, and thus the meaning that it carries in its context (why is it said at this particular point? What has just happened?) can be lost. This is particularly important in a study exploring the discursive
construction of identity, where talk is not being taken at face value but also for how it functions in the conversation. In order to combat this decontextualisation, I frequently returned to reading the interviews as a whole, which was made easier by the relatively small dataset. I also sometimes revisited the recordings when
interpreting particular sections of data, in an attempt to combat what Steinar Kvale has called the “violence” of transcription, where the nuances of tone are lost in translation to the page.
Listening for talk
I have already explored the situational context of the interview interaction, and some of the complications, limitations and possibilities of interpreting talk and relating it to experience. The interview as a genre, which carries particular conventions and expectations, is widespread and well-known in western society through media (Plummer 1999); both interviewer and interviewee10 are aware of and shaped by these conventions. So throughout analysis, and through the thesis, I take this context into account, not only through understanding and acknowledging limitations, but also through using the context as a source of data, interpreting how young people make use of the interview setting to perform gender, class, sexuality and age (Allen 2005b). I was looking for the “discursive positionings” (Davies and Harre 1990) taken up by participants within conversation; how they positioned themselves as particular kinds of subject through interaction with me, and, in group interviews, with the other participants. I wanted to explore what Walkerdine et al. refer to as “subjectification (the production of 'the subject' in discursive practices) as well as
10
While I use the term “participant” throughout the thesis, here I use “interviewee” to draw attention to interpretations of the setting based in prior experience, which cannot be wholly erased no matter how an interview is carried out.
98 subjectivity (the lived experience of being a subject)” (Walkerdine et al. 2001, p. 176) – a matter as much of what was not said (as I discuss shortly) as of what was. How did young people produce themselves in the context of the interview? What kind of narratives did they tell about themselves and about others?
Analysing the words that young people used, and the way they used them, I was reading for, as Holland and Ramazanoglu summarise, “ideas, beliefs, norms, discourses, reproduction of culture, and their effects” (2002), as they circulated around gender, sexuality and class. I sought to examine the underlying discourses of heteronormative masculinity and femininity, and how young people negotiated these in their talk about their own and others’ experiences. As I analysed the interviews I became increasingly focussed on the situation of subjects in their wider social contexts, and the ways in which subjectivity was being produced through interaction with and imagination of others. I use the term “relationship cultures” throughout the thesis, to illustrate the ways in which young people's (sexual) subjectivities were bound up with their friends and peers, and to highlight the extent to which young people's peer cultures were heterosexualised within school. Analysis of this
intersubjectivity was differently possible within group and individual interviews, and would, of course, have been different again through ethnographic observation. Listening for silence
Seeking out discourses, practices, and identity constructions is partially a matter of looking for patterns and connections, for how young people try and make themselves as coherent subjects. But, as I have discussed above, identity is not coherent. As Skeggs puts it, for the working class women she researched with, “searching for coherence is an impossibility, an ideal and a fantasy” (1997, p. 29). So it is just as important to seek out the contradictions, differences and tensions in young people’s accounts; the “discursive contradictions” that young people “live with and navigate... on an everyday basis” (Ringrose 2008, p. 41). Thus I paid particular attention
through the analysis to looking for equivocation, denials, disavowals, downplayings, and minimisations of emotions, statements and experiences. This is not to suggest that my interpretation of young people’s experience is superior to their own, but to
99 argue that not all positions are consciously known, willingly accepted, or freely shared (Hollway and Jefferson 2005), and that people represent their experiences in particular ways for particular reasons.
In a study on young sexuality and class, silences speak loudly. As mentioned in the previous chapter, class is often unmentionable, only referred to by proxies,
uncomfortable to speak about (Savage et al 2001). For young people, talk about sexuality, too, is often uncomfortable, and for me (as I have explored earlier) it was often difficult to navigate the lines of ethical practice, working out what was
acceptable to ask. The body, too, is central to understanding sexuality, but often unspoken. This is particularly so in relation to teenage girls, for whom desire as embodied feeling is often difficult to articulate (Tolman 2005, Fine 1988). It is, perhaps, especially difficult to understand young people’s relations to their bodies through the verbal method of the interview in the desexualised space of the middle- class school. But I agree with Sara McClelland’s argument that feminist researchers need to excavate and allow space for girls’ desire, “even when it is denied or
stuttered” (2008, p. 255, emphasis in original). As such, I have tried to read through to spaces for pleasure and desire, as well as looking for times and techniques through which desire is denied, disavowed, silenced, by girls themselves and by others.
Throughout my thesis, I have aimed to remain aware of the responsibilities of analysis: that is, as Holland and Ramazanoglu put it, “interpretation is a key point in the exercise of power” (2002, p. 116). Although I am critical (sometimes, in both senses of the word) of the young people I have studied, I aim to represent them fairly, appreciate their contradictions and multiplicities, and produce knowledge that may work towards improving a small part of the world.