For my assessment as part of this course, I am producing a critical journal, detailing my assumptions and realisations about women's studies research and about myself as a researcher. It will include some preparation work for classes and reflections on this preparation, as I discover new ways of researching.
NOTE: Please feel free to read this blog and comment freely. I am not presenting facts, this is a record of my thoughts and processes. However I will ask, given that this will form part of my assessed degree course, that you please seek my permission before citing any of this material yourself.
There are many different sources available which can be used in person-centred research, interviewing being just one of them. These can include:
- Official documents (including birth and death certificates, marriage and divorce licenses, medical certificates, criminal records, the census). These sources are generally readily available from local records offices and national archives.
- Private documents (including letters, diaries, autobiographies, e-mails). However isues were raised in terms of both the accessibility of these artefacts - they would either have to be gifted by relatives or friends of the person being researched, or by the person themselves - and the likely class and status of a person who has extensive and accessible private documents. If these have been made publicly available, the person being researched is likely to be high-status. Thus the use of private sources for research is a preselection of who has been recorded in this way.
- Mass Observation Archive: This can be useful for providing accessible information about people's everyday lives.
How do we find respondents?
Penny Summerfield outlined several ways of locating respondents in her article, including advertising in national magazines, however the most successful method was snowballing. But using snowballing to locate research respondents can be problematic. As the respondents will usually be friends or family members, or will know each other through a group or organisation, snowballed respondents are usually similarly diverse and do not form a representative sample in terms of ethnicity, age, sexuality or class.
However whilst snowballing may be the most overt about potential skewings of data, it is important to realise that all forms of selection produce skewed data. By the necessary process of selection a researcher is leaving someone out, thus skewing the data. For example, content analysis, which appears to be more representative and less skewed than snowballing, still skews its data, even though it is less overt about it. A researcher must be explicit about the nature of the skewing which occurs, always set clear parameters and be aware of the limitations of the research.
My Project
Researching feminist activist groups in York, UK, 1880-1960. I would like to explore the kinds of women's or feminist groups which were actively campaigning for women's liberation and emancipation and the sorts of activities they were involved in. I would interview respondents exploring their own memories of their involvement in feminist activism and their memories of their mothers', grandmothers' and other family members' involvement. I would like to explore these memories in relation to reflections of what the respondents think of feminist activism in York now and how it has changed over the years and to allow modern-day feminist activists to (hopefully) look back at our 'first-wave' sisters and discover what day-to-day activities were really like for these women.
The immediate issue which was raised about this project was the huge time span. The period covered witnessed many changes in society, government and culture and thus the concept of activism may have altered over the years. I should therefore set more realistic parameters for this research.
Locating respondents
I would seek respondents who were active in feminist groups during the period, or women who have memories of other family members' involvement. I acknowledge that it may be difficult to locate a representative sample of respondents for this project, given that the period I am researching is quite far in the past thus I may need to confront the 'problem of survival' (Summerfield, 49). My main method for locating respondents is likely to be snowballing, with an awareness of the data skewing issues for this method. A valuable starting point for this could be current feminist activist groups who have kept in touch with former members or with women who have been involved in feminist activism in the past. Given that this research is geographically specific, it is possible that respondents are either themselves current residents of York or may have family or friends who are local residents. Therefore advertising in local newspapers, shops, cafes, Post Offices and on local radio would be useful. Current local feminist groups' own historical sources might also prove useful.
I was encouraged to begin to look at local archives, such as the Feminist Archive North and council archives for local organisations. Another valuable resource was pointed out to me - residential homes and sheltered housing. However the ethical and practical considerations for this were stressed. In order to interview residents, I would need to contact a manager. Whilst I first considered that this would be most appropriate in writing, it was suggested that this could be ineffective without any previous personal contact: people are much mroe likely to respond if they are approached directly. Thus a personal face to face visit or a phone call may be more appropriate.
Transcribing
Most researchers end up working from a transcript of their interviews, as this is generally easier. However doing this makes this research into a form of textual analysis. Some critics, notably Alessandro Portelli have warned of the danger of losing the interviewee's 'voice' in the transcript by ignoring nuances of tone, inflection and volume in the recorded voice (Summerfield, 51). Sometimes issues of translation cause problems, as there is not always a direct translation from one language to another.
The linguistic conventions in the way we talk are culturaly specific - certain types of narratives produce certain stories and follow a particular narrative structure. Thus there are underlying assumptions which a researcher and a respondent will make about the interview and the questions an interviewer asks can influence the respondent's answers. In selecting a criterion on which to base a research project, you are asking respondents to rearrange how they form their lives around this criterion, for example sexuality. This may give a false dominance to this criterion or this aspect of their lives as it may not be the way they would interpret their own lives.
The interaction between interviewer and interviewee is also significant in that age and outward appearance can make a profound difference to the dynamic of the interview and thus to the interviewee's responses. Even in a text-based interview, for example via e-mail or through a chat room, your positionality is still visible in the language you use and don't use (slang, technical terms, abbreviations).
Interviewers can project their own norms through their questioning, forcing people into telling a certain story.So researchers must seek neutrality as much as possible, so as not to lead the respondent astray or to restructure their response and mould it into something it is not. However the source materials themselves also produce norms, in that they are informed by normativity and this affects the stories they can tell. There are multi-layers of norms which affect person-centred research and these norms of narrative will be unspoken and implicit.
I will try to identify feminist groups which were active at the time through various sources including (but not limited to):
- newspaper archives
- Council archives
- local archives (the Feminist Archive North)
- current feminist groups' own source materials
I will use snowballing techniques for my research, beginning with current local feminist organisations, personal contacts and further contacts through my first few respondents (hopefully). I will also advertise this research in Post Offices, shops and on local radio. I will use the valuable resource of residential care homes and will approach these directly to seek respondents.
Hopefully, this project would be completed within nine months to a year, so I would estimate that I could conduct 10-15 interviews within this period, with suitable time for writing up and analysing data.
If anyone would be interested in this project I think it is a valuable one which I would be very happy to do!