This post is written by Dr Peter Matthews who works in the School of the Built Environment at Herriott Watt. Peter’s blog is Urbanity…History and he tweets as @urbaneprofessor. I asked him to show and tell how he talked about the messy bits of his research because I know it’s something that doctoral researchers are anxious about. How much to say, how open to be? Check Peter out when you’ve read what he has to say about writing his thesis.

Following Pat’s recent blog post on messiness in research we conducted a wee discussion over comments and then emails about a particular approach I took to my doctoral research (published variously here and here with yet more to come). The methodology was heavily informed by interpretive policy analysis – specifically the wonderful text about Israeli Community Centers by Dvora Yanow, How Does a Policy Mean? (Yanow, 1996).

In Dvora’s extensive writing on methodology she emphasises the role of writing in analysis. In writing up my thesis I was also taken by the idea of methodological story – discussed by Laurel Richardson in her chapter for the wonderful Sage Handbook of Qualitative Methods (Richardson, 2000). Like any chapter in a thesis, this took a lot drafting and re-writing. According to my memory stick the version that is actually in the thesis, collecting e-dust on the electronic shelves of the Glasgow University Library, is the fourth, but I remember there were more versions than that. The aim of the chapter was to completely rewrite my “normal” methodology chapter, put the “I” back in, and be explicit about all that messiness of the research process.

This worked because the whole thesis was appealing to an ethnographic standard of validity – its thoroughness and explicitness about the methodology testified to the degree of immersion in the policy environment I was studying. Pat’s post and further prompting got me to look back at the chapter and she invited me to post here. So I’m going to provide two excerpts, both revealing this messiness, but with different impacts on the validity of the research as a whole. I’ve had to change them slightly to preserve anonymity and so they make sense, but they do give an honest insight into the messiness of the research process and raise important theoretical and methodological questions.

A positive story

To provide some context for the narrative – my doctoral fieldwork lasted around a year in 2007-8 in two local authorities in Scotland. I have anonymised the places to protect the participants. I was mixing methods of unstructured narrative interviewing (where the intention is to get your participants to tell you an unrehearsed story) and overt non-participant observation (sitting in meetings, oh so many meetings, taking notes). Participants and meetings ranged from community activists and the meetings of local tenants’ and residents’ organisations to strategic officers in the local authorities and the meetings of strategic partnerships – community planning partnerships.

In this excerpt from my methodological narrative I write honestly about the positive experiences of using the method:

“Narrative interviewing was specifically chosen for its high standard of ecological validity (Hollway & Jefferson, 2000). The research was interested in the meanings derived directly from interactions with the policy process (Yanow, 1996), not those created when participants “performed” for interview (Fontana & Frey, 2000). Narrative interviewing aimed to achieve this by enabling free association through the production of narratives derived from the open-ended questions (Hollway and Jefferson, 2000). Some interviews clearly achieved this. When I interviewed community volunteers in their homes they would be sat in their favourite chair, cigarette in one hand, cup of tea in the other, and just talk. It felt like someone was asking them about something that had meant a lot to them, and this was the first time someone had taken them out of their everyday life back to this important time for a while. These were narratives that had been shared and discussed in the kinship and friendship networks of the community throughout their time as a community volunteer and resident. Through them, they understood policy; what had happened to their community during their residence; and what their role was in this. In these interviews with community volunteers, I noticed the importance of setting, especially those carried out in formal meeting rooms. These participants were initially worried they were saying the “right” thing. I had to reassure them that I just wanted them to talk and there was no “right” thing to say. After a while they did relax and just talked, but it was more difficult than when they were in their own home.

Strategic officers and community development workers struggled with the interview technique more and needed to be prompted a great deal, using the cues in the interview schedules. Overall, interviews with officers tended to last around 45 minutes, compared to one hour, and up to three hours, with community volunteers. In many interviews the narrative interview technique failed with strategic officers. This does question the cohesiveness of the methodology and the validity of the data, especially whether the data from strategic officers is comparable with the data from community volunteers. However, although a free-flowing narrative was not produced, the interviews definitely did become negotiated texts (Fontana and Frey, 2000). In one interview with a strategic officer we reached the end of the interview schedule in less than half an hour, just as the participant was getting into their stride, relaxing and providing a wealth of valuable data. I stumbled over my words, explaining that I had come to the end of my questions. The participant retorted that ‘I’m just getting into my stride now‘. I closed the file that contained my interview schedule and said: ‘I know that’s what I was thinking as well do you want to start again?‘ For the last 15 minutes of the interview we just talked at ease about what regeneration policy and Community Planning were trying to achieve. Further reflection on these “failed” interviews (Nairn, Munro et al., 2005) in the process of analysis also showed that this “failure” actually provided additional data; for example when a strategic officer struggled to explain the role of the community in the policy process they actually revealed further evidence of major problems in community engagement in policy processes.

Again, setting mattered a great deal. I only really achieved the same level of engagement with professionals as I did with community volunteers when I was in their office – their comfortable territory. A lot of the interviews with staff from the [local authority 1] were carried out in the atrium of their new headquarters. This was a large, open-plan, noisy space. I knew from my own experience that the building had been purposefully built with little meeting space for efficiency reasons, with a notion to use the large amount of public space as an area for informal meetings. Many participants therefore seemed to think they were in the space for a formal meeting and took on the formal role of an “interviewee” rather than themselves. They would comment that some things they said were ‘off the record’ and should not be attributed to them, whereas those in their own offices did not share this concern. It was striking what a difference a cup of coffee could make. Two participants got themselves cups of coffee, and it could have just been their personality, but they seemed a lot more comfortable with their hands wrapped around the cup, talking over it, taking sips as they gathered their thoughts.

By the end of the research process, the on-going process of analysis (Coffey & Atkinson, 1996) meant I was bringing findings back into the interview discussions with participants – primarily at this stage, strategic officers and local councillors. This provided a wealth of data about the very basic questions of regeneration and Community Planning and what they were trying to achieve. This revealed how they negotiated a role for themselves within the policy context; what their job meant to them and the criticisms of Scottish and local government policy they had. “

In my next post I reveal a less positive story, questioning the validity of my findings based on this methodology and reflect on the use of the narrative in my thesis and research and the implication of this approach to writing up ethnography for a wider audience.