|
Chapter Three The methods of data collection and analysis I have chosen for this research incorporate both qualitative and quantitative approaches. This involves participant observation, conversations, structured interviews, and questionnaires. As my research is based upon online communities I have chosen one specific community as my case study; therefore all data collection will take place within this community. I believe a case study will provide data through in-depth investigation and participation. However, taking a case study means that analysis is time and space specific. The case study I am using is ‘Caithness Community Web Site’. This research involves interpretive approaches and sequential analysis. The research focus is whether theories of community networking can be applied to enhance social capital. Interpretative reading of data involves constructing or documenting a version of what the researcher thinks the data mean or represent, or what the researcher can infer from them (Mason, 1996:19). In this research qualitative and quantitative methods are utilised in order to improve validity and reliability of data. This has both positive and negative implications. ‘Multiple methods give a fuller picture and address many different aspects of phenomena, however multiple sources of data demands multiple data analysis skills’ (Silverman, 2000:50). The flexibility and sequential nature of research in the social sciences allow research design to incorporate both quantitative and qualitative approaches (Mason, 1996:34). Participant observation allows the researcher to discover the every day practices of subjects and provides an over-view of the community from an insiders perspective. Problems with participant observation stem from the reliance of opportunistic moments, symbolic interaction, and trust. Chatting to people in a virtual community involves an element of trust on both sides. Also non-verbal communication is only through ‘smilies’ - textual expressions. Interviews allow the researcher to ask questions specific to the study. These can take the form of conversations, semi-structured interview questions, or formal, structured interviews. ‘Interviewing provides a way of generating empirical
data about the social world by asking people to talk about their lives. In
this respect, interviews are special kinds of conversation’ The interviews are coded using microanalysis of data. By analysing frequently occurring words and themes categories are generated and applied to theories outlined in the literature. Clarification of the material involves eliminating superfluous material such as digressions and repetitions, distinguishing between the essential and nonessential. Literal diagrams are used allowing connections and relationships within the data to be analysed quickly and easily, this is in the form of a flow chart of themes and arguments. The questionnaire contains closed questions providing useful information for the study, but it does have limitations. Firstly, the sample consists of voluntary respondents. Through this method the researcher has no control over the number of responses, which could be zero. Also, this method cannot ensure a sample will be representative of gender. A self-completion questionnaire follows a standardised format in which most of the questions are pre-coded and provide a list of responses for selection by the respondent (Gilbert, 1993:96). The closed questions are pre-coded but provide limited answers. This is problematic as some respondents did not answer all questions, leaving incomplete information. Finally, the questionnaires are anonymous which means the respondents can not be contacted for further questioning. Both quantitative and qualitative methods are applied via email, this itself has limitations. Timing is asynchronous, therefore the researcher cannot prompt or explain any concepts or questions. Also, the researcher cannot asses non-verbal communication such as hesitations or body language. Email is impersonal and does not allow a natural conversation to develop, hence respondents have to be contacted repeatedly. Next, quantitative and qualitative methods used in this research will be explained independently. We will begin by outlining the quantitative methods and why they were chosen.
I decided on quantitative questionnaires to counteract the disadvantages of participant observation - limited contact and reliance of opportunistic moments. Sampling for the questionnaire is from my case study; all users within this site are targeted. The questionnaires have been designed to provide an overview of who uses this Community Network. This would not be a representative survey of community network users as it specifically targets one group of people. The questionnaires consist of closed questions in order to gain precise answers. Questions were chosen to coincide with the information given on the Community Networks ‘home-page’, examining what aspects of the network people are most interested in. Also, how often these community members contact the site and where they log on. Questions about nationality, age and sex also give a broader insight of the users. I decided on one open question at the end of the questionnaire asking users to rate the site. This proved very interesting and provides personal reasons for visiting the site as well as general opinions. The questions I need to ask about trust, community, and networking I feel are more appropriate within qualitative techniques. Therefore, these are asked through conversations on the chat page and over email. I feel my target population, users who are interested in visiting the site for recreation, may not understand questions on social capital resulting in a lower response rate. Therefore I chose basic, straightforward questions on users personal details and specific interests. Distribution of the questionnaires was via email. The questionnaire was emailed to the community networks technical adviser who posted it onto the home-page. Therefore respondents were volunteers, I found this helpful as most people included extensive comments in the open-ended question. The questionnaire was posted on the home page for three weeks, allowing both frequent and occasional users to participate. The completed questionnaires were secure information and emailed directly to myself. In total I gathered 82 completed questionnaires. If I could have done anything differently I would have differentiated between British and Scottish as nationalities, I suspect many Scottish people would state their nationality as British. I would have also included a question on users occupation, as I suspect the majority are professionals. A copy of this survey can be found in appendix one. We will now examine the qualitative methods adopted in data collection and analysis.
As the aim of my study is to investigate social capital within a Community Network my research takes place online. Participant observation is achieved by becoming a community member; this is achieved by gaining a user name and password allowing me to participate at any time. Caithness community web site offers both a message board and chat room in which asynchronous and synchronous communication can take place. Site facilitators moderate these interactive pages; bad language and derogatory comments are strictly prohibited. My methods are overt throughout, stating who I am and what I’m interested in. I openly invite people to chat about their reasons for becoming a community member and their feelings toward it. This way I have gained new contacts and an insight into the feelings of collective solidarity the site provides. This co-operative method of inquiry helps to generate ideas from others and incorporate them into the research process. ‘Persons can only properly study persons when they are in active relationships with each other, where the behavior being researched is self-generated by the researchers in a context of co-operation’ (Reason, 1994:41). During primary contact with the sites facilitator’s many conversations about the reasoning and theories behind the project provided interesting views which have become incorporated into the data. These led to the design of structured interview questions addressing key issues. Microanalysis of this data generates categories from the language used in interviews and conversations. These categories include, ‘community’, ‘reputation’, ‘local people’, ‘funding’, and ‘future visions’. The structured interviews provide open-ended questions for a small sample. My sample was taken of the sites voluntary facilitators, or ‘webmasters’. This proved problematic as the four webmasters involved in this project are all male, therefore the sample was not representative of gender. However, because the sample was so small I had 100% response rate. My original research plan included conducting interviews with webmasters from other community networks, however, after just one interview with another community web site it was decided this was too time consuming and unnecessary. Focusing on Caithness I am interested in finding out what kind of people facilitate a community network, what is involved, and why people volunteer for such a position. My questions included personal opinions about the role the site has in the local community, and individual goals webmasters hoped to achieve. This was interesting as, although the site has common aims and objectives, individuals have personal objectives for the future of the site. I also asked if they related their work to ideas on social capital. This question proved problematic as different people interpret social capital in different ways, although it did offer many diverse answers. A copy of the interview questions can be found in appendix two. The interviews were completed in the respondents own time via email. The main problem with posting the interviews online was that I could not probe interviewee’s for further information, also they could not ask me for explanations to questions they did not understand. The final question was if they would be willing to be contacted again, all replied positively, therefore I used these qualitative interviews to further my investigation into the feelings, ideas, and beliefs behind community networks through informal conversations via email. We shall now move onto look at the actual analysis of the data. |