Chapter 3 describes the methodology I employed to set about answering my Research Questions I justify my choice of a qualitative, ethnographic approach to studying CCPs, and I discuss some of the issues that might
2) My own research methods, including ethics of research, the types of data that I can collect, what these data represent, and how to handle and analyse these data.
3.2 Proceeding with field work
3.2.1 A note on research ethics
In social research, ethics has been an important component of designing, carrying out and writing-up fieldwork projects for many decades. It has been widely-recognised that studying the cultures and societies in which individuals are embedded has the potential to cause significant emotional harm to these individuals and societies. For example, the researcher may use methods that those studied find intrusive, or may produce accounts of their fieldwork that result in substantial societal disruption or exposure of individuals. As a result, there has been much of consideration about how social researchers should proceed, drawing on a variety of ethical theories and how they might apply in different situations (for instance, Ess 2006).
Translating these approaches into an ethical framework was not a straightforward task. I chose not to develop a rigid framework before entering the field, for two main reasons. The first is that, given the relative newness of social research on the Internet and the difficulties inherent in transferring offline research methods to online contexts, there is a marked lack of consensus regarding how to develop such a framework (Buchanan 2011). The second is that, given my methodological commitment to try and understand the communities I was observing in their own terms, I did not want to pre-judge what would be appropriate behaviour as a researcher. Instead, following Markham’s (2006) call for a ‘reflexive’ approach to online social research (p. 37), I entered the field with a set of sensibilities that I would reflect upon and refine, and how these were applied to my research will be discussed in subsequent subsections.
These sensibilities were informed by existing literature on the ethics of internet research (a particular influence here was the report of the Ethics Committee of the Association of Internet Researchers, which is widely- cited (Ess & AoIR 2002)). I was especially sensitive to the notion that these online communities were potentially very meaningful for the participants (both professional scientists and software engineers and volunteers). From the perspective of the scientists and software engineers, most had a high professional stake in their project, and I was therefore concerned that my research would not prove disruptive to this.
From the perspective of the volunteers, I became aware of the importance of the online community to participants in terms of socialising and online relationships and sometimes the opportunity to construct identities in ways that were not necessarily available off-line (for instance, one disabled participant explained to me that
56 nobody in the forum knew they were disabled and they valued not being treated differently from others as a result of this). It also became apparent that the online interactions were important to subjects in their offline lives as well. For instance, one volunteer on Galaxy Zoo organised a spinoff research project using threads on the project’s forums, and has now used this as a basis for starting a PhD at the Open University, whilst another has been encouraged to pursue science communication as a career, encouraged by responses to their forum posts discussing the project science. These ethical concerns had a number of implications for how I approached my research at all stages. These will be discussed where appropriate in the following subsections.
3.2.2 Interviews
Interviews comprised a very important part of my data. During the course of my research, I conducted interviews with 22 scientists and software engineers and 36 volunteers across the two projects. These interviews were conducted using a variety of media: face-to-face, Skype, Instant Messaging, and email. All interviews were semi- structured (Arksey & Knight 1999). What this means is that a schedule of questions was drawn up in advance of the interviews, however I was able to deviate from that list during an interview, for instance by asking questions out of order, and following up interesting and unanticipated lines of inquiry that arose. Such an approach combines the benefits of structured interviews where a list of questions is strictly adhered to (and thus allows for comparisons to be made across interviews) with unstructured interviews where there is no schedule of questions (which gives the researcher the flexibility to follow-up unanticipated lines of inquiry).
My ethical concerns guided how I proceeded with recruiting and conducting the interviews. I set out to secure informed consent from those I interviewed to allow me to make a digital audio recording of the interview (where it was face-to-face or conducted via Skype) or to keep a transcript of the interview (when conducted using Instant Messaging or as a series of emails), and to use quotes from the interviews10. The privacy of interview subjects was protected by keeping all of my data (both digital recordings and transcriptions) in a password- protected area of my computer, and storing any printed transcriptions in a locked filing cabinet. Only I have seen these transcriptions, or listened to the interview recordings.
10 See Thorne (1980) for some of the issues involved in gaining consent. Varnhagen et al. (2005) discuss particular difficulties in
making sure that informed consent truly is informed in the context of online research, for instance that the fewer visual and audio cues in mediated communication (compared with face-to-face communication) can complicate the researcher’s task of ensuring that understandings between researcher and interviewee are reached. I made an effort to take these concerns into account.
57 I recruited interview subjects in different ways according to whether they were volunteers or project scientists. I sent emails to every member of the team that I could identify (past and present), as well as to scientists who were not part of a project’s core team but who had collaborated with members of the project on work relating to the project. I sent individual emails to each, in which I introduced myself, briefly introduced my research, and attached an Information Sheet giving more details about my research such as its motivations, methods and expected. I asked whether they would consider being interviewed by me for my project. If they replied in the positive, I then sent further emails to establish a time and date for the interview, and I asked them explicitly whether I could record the interview and use (anonymized) quotations and also whether they had any further questions about my research. I sought to conduct as many interviews face-to-face as possible, although this was not feasible for many who were located abroad. In the event, I conducted 22 interviews with Galaxy Zoo and climateprediction.net scientists, either face-to-face or via Skype. All interviews lasted between 30 minutes and 90 minutes, with the majority lasting 45-60 minutes.
In the case of volunteers, I faced one challenge in particular with recruitment, namely that those running the project were not comfortable with me contacting volunteers directly to ask whether they would take part in an interview. In order to avoid spamming volunteers with unsolicited requests, I instead devised a short online survey to recruit potential interviewees, which I promoted on the project forums with the support of forum moderators. Two drawbacks of recruiting using this method are, firstly, that it does not guarantee a representative sample of volunteers as participants are self-selecting, and secondly that it restricted interviewees to those who actively consulted the forum. This excluded those who, for whatever reason, chose not to participate in the forums or who no longer participate in the project. Given the circumstances, however, this form of recruitment was the best that could be achieved. Furthermore, given that I had undertaken to study the community that formed around a project, it did not seem to be too much of a problem not to speak with those who did not actively constitute this community.
This survey asked eight brief questions about the volunteer’s involvement with the project, and also consisted of a box at the bottom where the volunteer could indicate if they wished to participate in an interview. As with my interviews with the project scientists, I sent an information sheet and an explanation of how interview data would be stored and used and a request for consent.