Fire Fighters, Neighbourhoods and Social Identity: the relationship between the fire service and residents in Bristol

Writing as research method

As discussed above, Van Maanen (1988) suggests three different types of writing for ethnography: realist tales, confessional tales (where the writer is seen as an instrument of the research) and impressionist tales, a more post modern take on social science writing which gives voice to some of the research participants using a variety of literary styles. In the realist typology, writing is presented as a finished product, not something for either the reader or the writer particularly to engage in or with (Van Maanen 1988). This might be particularly characterised with earlier anthropological or sociological texts, which lack the reflexivity more usual in modern qualitative research.

Richardson (2004) looks at the subversion of the writing process undertaken in using it as a method of analysis. She reflects that when we are taught to write, we are taught not to write until we know what we want to say. By taking Van Maanen’s style of ‘confessional tale’, she now writes ‘in order to learn something that I didn’t know before I wrote it‘. Colyar (2009) develops this argument further, viewing writing not just as method, but also as learning in and of itself. For her, the self reflexivity of the process is its strength, ‘because the writing process generates a text, one we can use to better understand our rhetorical selves, writing seems an obvious means of pursuing self reflexivity and therefore more trustworthy research findings’ (page 432). As such, we come not just to know about our work but about ourselves. Typical of this ‘moment’ of ethnography (Coffey 1999) Colyar’s work is not presented as a standard academic text, rather in a less formal, contrived ‘work in progress style’ (Colyar 2009). This clearly reflects Van Maanen’s third type, as does ground breaking work such as that presented by Paget which takes interview transcripts, and relays it as performance (Paget 1990).

The writing process has been integral to this research project, and has remained consistent throughout the three studies, with the additional use of vignettes – an explicitly reflexive practice (Richardson, 2000) – in the third study. Fieldnotes, observations and interview transcripts formed the majority of data in each study. This was then collated into qualitative analysis software. Although conducted ostensibly through thematic analysis (Braun, 2006), writing was an integral part of the analytical process, especially in Study Three, where I have used vignettes to describe observations of the community. The use of vignettes encourages a greater level of engagement with the data for the reader, asking them to relive the experience through the eyes of the researcher (Denzin 2000). This has the effect of enhancing the contextual depth of the material (Miles, Huberman 1999), ‘bringing life to research and research to life’ (Humphreys 2005). As with Humphreys, my vignettes are derived from fieldnotes taken contemporaneously, and comprise a combination of reporting of events with a degree of introspection.

Analytic strategy

Drury and Stott (2001) advocate the development of a ‘consensual’ (page 365) account of events, which relates to methods of triangulation proposed by Denzin (2000). This suggests that an account can be recorded between participants which is based on agreement between parties. Although this does not attempt to present an absolute truth, and remains necessarily constructed, it does serve to present ‘the reality as understood by the various parties’. However, such an approach presents a problem for the lone researcher, as there is no one with whom to compare interpretations of events, and so the development of a consensual account is problematic. Nevertheless, individual accounts can still maintain a degree of validity (Flick 2008) with themes and accounts ‘crystallising’ (Richardson 2003) around certain factors, although it remains to be said than any interpretation and reconstruction of events is mine alone.

As with Drury and Stott (2001), these data were gathered ‘opportunistically, rather than systematically’ (page 366) and, as such, are not intended for quantitative analysis. They are indicative of my interpretations of comments made to me, or within my hearing, by a small number of people, who, although potentially typical of AFRS fire fighters in particular and personnel in general, do not claim to speak for the entire organisation. Likewise, whilst residents might speak as a resident of a particular neighbourhood, they are unlikely to have claimed to be speaking on behalf of it.

Having recorded and annotated fieldnotes and collated them, the analysis itself was conducted using thematic analysis, a qualitative method which seeks, identifies and analyses patterns within data (Braun, Clarke 2006). Braun and Clarke describe thematic analysis as a ‘flexible and useful’ means of analysing data, which is capable of producing ‘rich and detailed, yet complex’ interpretations of the data (page 78). It enables the researcher to take an active part in the exploration of the data, rather than relying on ‘emergent’ themes or the commitment to generating theory inherent in grounded theory approaches (page 81). As it is not my intention to produce further theories with my research, rather to expand on existing ones, this is a useful distinction.

The aims of this research programme are to understand better the relationship between AFRS and the neighbourhoods in which they work, by looking at the nature of the relationship, examining the roots of hostility and resistance and asking whether current engagement mechanisms are effective. As such, the data are approached with these questions in mind. In each study, the analysis was conducted by coding into initial themes, based on recurrent or particularly prominent ideas, then reviewing and recoding iteratively as further themes emerged, a process which bears similarities to other types of qualitative analysis, as described by Brunsden and Hill (2009). As such, the coding schedule did not reach completion until the coding process itself was close to completion. Further analysis was conducted as the themes were written up (Richardson 2003), and as such, the writing has been as much of the process of inquiry as its reporting. As with any research, it has been necessary to strike a balance between a rich account of the full data set (Braun, Clarke 2006) and a nuanced account of particular themes (page 83). This is especially evident in the focus group study, where I have attempted to provide a rich overview of the majority of themes, whilst also providing a more detailed analysis of some.

Download 0.87 Mb.

Share with your friends: