3.1. The Planned Ethical Approach to Confidentiality
The terms confidentiality and anonymity are frequently confused (Lahman et al. 2015
), so it is helpful to provide contextual meaning. Confidentiality is presented by Walford
(2005, p. 85
) to mean information that is private or secret that should not be passed onto others. He argues that, although researchers should keep confidential any information that participants ask to remain off the record, there can be no blanket confidentiality, as the purpose of conducting research is to gather information to analyse and share with others. The proposed solution is therefore to ensure that confidential information may not be attributed to a particular person. Anonymity, which may be more correctly termed pseudonymity, is described by Wiles
(2013, p. 50
) as a key method applied by the researcher to protect the confidentiality of research participants “through the use of pseudonyms applied to research participants, organisations and locations”. This “ubiquitous” practice is seen as the “desirable standard” or “default setting” in order to achieve the minimum requirement in research of “doing no harm” to participants (Guenther 2009
; Tilley and Woodthorpe 2011, p. 199
; Walford 2005
). It is therefore unsurprising that the use of pseudonyms for all participants was designed into this research without considering alternatives.
Ensuring that research is undertaken ethically has moved from being a background assumption to an overt aspect of research planning and practice (Brydon 2006
). As part of the overall governance of this research, ethical clearance was sought in advance of field work and granted by the Social Sciences Departmental Ethics Committee of the overseeing English university. This scrutinous process required evidence that each potential participant would be equipped to make an informed choice on whether or not to participate. Information developed in response to that requirement covered key areas such as: the purpose of the study; the approaches to confidentiality and anonymity; and assurance that participants could withdraw consent at any stage of the process. This information was sent to participants in advance of each interview (with the exception of two participants in Namibia recruited through intermediaries) and always discussed face to face before the interview. Participants were informed that their interview recordings would be stored securely until the point of being destroyed and would not be shared with anyone else, and that their identities would be anonymised, as would any third-party associates discussed in interviews. Guenther
) argues the importance of researchers explaining their decision-making around how pseudonyms are used. In this instance, the author asked each participant to choose her own pseudonym before the interview and explained that she, as researcher, would obscure the identity of third parties after the interview, in line with ethical requirements. The rational for asking participants to choose their own pseudonyms was to enable each to read her content in the final research produced, should she wish.
3.2. Research Participants’ Rejection of Pseudonymity
During the early stages of the first period of field work in Namibia, three participants expressed uncertainty about pseudonym use. The author explained it was an ethical requirement of the study to ensure that participants would be protected from harm and the participants consented/chose their pseudonyms. It was an unexpected development when two participants interviewed towards the end of field work in Namibia refused to assume a pseudonym during pre-interview discussions. Rosa and Sarry separately and individually stated that their life histories could be used on the condition that they were ascribed their own names. The author proceeded with both interviews, though explained that it could not be guaranteed that they would be used in the final research, as new ethical clearance would be required, which could be refused.
Each participant provided her reasons for refusing an inauthentic name. Rosa explained that she has publically campaigned on the issue of VAWG in Namibia for many years, even before independence, and had been subjected to long-term mistreatment because of it. She queried why she should use a false name and stated she was not willing to do so. Rosa added that, as an activist well-known throughout Namibia, people would be likely able to identify her, regardless of the name given (a factor already under the researcher’s consideration). Weeks later when Sarry refused to assume a pseudonym, she explained that a key aspect of her activism has been supporting women to publically share their experience of being violated in order to challenge stigma and victim blaming, particularly in relation to sexual assault. This included supporting women to speak at events and arranging for their testimonies to be written as a matter of record. She concluded that she just would not feel comfortable using another person’s name and posed a challenging question: “how can you write about a person who does not exist?” Each activist left the researcher very clear that her sense of identity was intertwined with ownership of her own name and that “anonymity implies stealing someone’s identity and history” (Svalastog and Eriksson 2010, p. 109
Rosa and Sarry’s insistence on being present in the research in their own names, and the uncertainty expressed by other participants, are understandable given the public stance many activists take on VAWG in their own names. Campaigning to change hearts and minds and advocate for policy and resources to change the status quo in relation to VAWG is a key aspect of feminist activism (Sweetman 2013
), and in that context one can envisage how using a pseudonym would feel counter-intuitive. While not all activists interviewed in Namibia did so, both Rosa and Sarry identified as feminist. Yoder et al.
) argue that those who describe themselves as feminist are likely to see the connection between the personal and the political, which can influence their engagement in activism and their wish to be identified as an activist. Furthermore, research undertaken for the Association of Women in Development (AWID) highlights that women’s rights defenders across the globe have been threatened, ridiculed, and violated as punishment for speaking out (Barcia 2014
). Against this harsh reality, individual VAWG activists such as Rosa, who spoke of such experiences, may perceive the risk of being named in academic research as remote and irrelevant.
Researchers have responded to participants wishing to be named in research in various ways. Guenther
) did not accommodate participants’ requests in her research on women’s activism, based on a concern that they might not foresee the harm that may result from being named, although this left her with an ongoing unease. Having been refused ethical clearance to name research participants, Barratt and Lenton
) argue the need for increased flexibility to renegotiate ethics procedures on pseudonyms, although they acknowledge that waiving anonymity may be problematic in relation to ‘higher risk topics’ (Barratt and Lenton 2010, p. 74
). Given the aforementioned challenges faced by women’s rights defenders, VAWG activism has the potential to be a higher risk topic, although the extent of this cannot be foreseen in advance of an interview. Following one participant’s distress at seeing the pseudonym in place, Grinyer
) offered participants a choice between using their own names or pseudonyms, an approach of relevance here.
The author wanted to accommodate Rosa, Sarry, and other participants’ presence as research participants under their own names, seeing this as an acknowledgement of their agency, and thus pursued a mixed naming approach as practiced by Grinyer. Feminist scholars who employ a participative research practice include Birch and Miller
), who challenge researchers to be more experimental in order to increase participant involvement. The author’s application to the overseeing ethics committee to accommodate a mix of pseudonyms and actual names was approved, subject to the implementation of the proposals contained in the application. These were that all participants, including those previously interviewed, be offered choice on the use of name, although anonymity would be recommended to all who did not hold views against it, and that obscuring the details of third parties discussed in interviews would continue as before.
At the time of writing, a total of five out of the twenty participants in Namibia and Northern Ireland have chosen to use their own first names, a number which is subject to change before the thesis is concluded. It is worth highlighting that not all Namibian participants have yet been given the option to review name choice, as the author hopes to pursue the follow-up in person so as to overcome challenges such as language constraints and lack of email access. Another noteworthy point is that, as ethical clearance was granted before field work in Northern Ireland commenced, all participants there were given a choice on how to be named from the outset.
Names have ethnic and other identity-based connotations which can impact research in complex ways (Lahman et al. 2015
; Wiles 2013
). It is interesting to note that one activist in Northern Ireland used this research as an opportunity to revert back to her own name, which she does not use in the context of her activism. As a name that strongly indicated her ethno-nationality/religion, she amended it years earlier during the conflict to make that aspect of her identity less obvious to others. Another participant chose a pseudonym which obscured her ethno-national identity, or rather suggested she was from the ‘other’ main group in Northern Ireland. In order to minimise the risk of confusion and misunderstanding that may accompany a mixed approach, a simple distinction is that names in italics are pseudonyms and names in roman text are the participants’ own.
3.3. Dilemmas in Attributing Data to Participants and the Informed Consent Process
Every datum attributed to a participant in written research, such as a quote or anecdote, influences how that participant is portrayed. This portrayal can make a participant identifiable even with a pseudonym in place, thereby demonstrating that confidentiality extends beyond anonymity/pseudonymity. Extensive literature points to the complexities of effectively obscuring participants’ identities including the consequences of being unsuccessful in this task (Lahman et al. 2015
; Scheper-Hughes 2000
; Snyder 2002
; Walford 2005
; Wiles 2013
). Even in research that acknowledges participants, a researcher is still required to reflect on how participants are portrayed, particularly in relation to sensitive disclosures. Regular examples of sensitive data from these life history interviews include intimate details of participants’ lives, participants being critical of others, and off the record disclosures, such as past actions with which the participant does not want to be associated.
One particular example features here, chosen because the participant’s response to her own disclosure brought into focus her lack of understanding as to how her life history data would be used. The participant, who originally expressed that she did not wish to use a pseudonym, but later changed her mind, made a disclosure about her past that she said would cause huge levels of distress to an immediate family member if discovered. During and at the end of the interview she asked the researcher not to share the details of the disclosure next to her name. Although the author was able to give this reassurance immediately, the participant asked several questions about how this would be done in reality. Her specific concern appeared to be that, as the disclosure related to an event which changed her life course, leaving it out might make the remaining story difficult to understand. It was at this point that the author realised that the participant expected the life history she provided to be retold in the same format in the research. This realisation caused the author to question whether other participants might have similar expectations.
Guidance on what to include as part of the life history informed consent process (see as an example della Porta
), who provides an indicative list of life history informed consent topics with activists) does not routinely include specifying how the data will be analysed and presented, which appears to be a gap. Although the author became sensitised to this methodological issue because of one participant, Smythe and Murray
) claim that it is a common misconception among life history research participants that the researcher will share the life story as provided by the participant. Wiles
) notes that there are varying views among researchers about the extent to which informed consent should include the ways data are used and points out that these ways are not always clear at the time of participant recruitment. In response, it is argued here that the locus of informed consent should also not be at the participant recruitment stage. Just as researchers’ knowledge about data use builds throughout the project, so too should participant consent.
Whilst acknowledging that researchers would not necessarily know the various possible uses of data at the recruitment stage, they would be expected to know the method of data analysis that will follow data collection. It is important that participants learn of this key method during the pre-interview informed consent process. Doing so will enable them to better comprehend what participation may entail and make informed decisions about whether or not to initially participate. This encounter prompted the author to ensure that subsequent informed consent discussions with participants included explaining that, although the interview comprises an end to end biography, the final research produced will be different in layout. The author explained that the research focus will be to discuss themes based on all participants’ testimonies, rather than including each individual participant’s life history in full. Participants expressed their understanding of this concept. Smythe and Murray
(2000, p. 331
) promote this type of transparency, though they suggest that informed consent could include explaining to participants that, just as journalists put their own spin on peoples’ stories, so too may a researcher. The author did not choose a comparison with journalism but did seek to be transparent with participants about the method of analysis.
Concern about the risk of portraying participants in ways that might cause distress to them brought the author to literature about ‘process consent’ or ‘iterative consent’ (Ellis 2007
; Lahman et al. 2015
; Smythe and Murray 2000
). Process consent is explained as an informed consent process, initiated by the researcher that is mutually negotiated with each participant, throughout the course of the research, including the process of data analysis (Smythe and Murray 2000, p. 320
). In other words, process consent means that the researcher checks out every datum with every participant and pulls any content that does not meet with the participant’s consent. Birch and Miller
) reflect that participatory practice, such as engaging participants in data interpretation, is difficult to maintain throughout research projects. Following field work, the author gave consideration to adopting process consent for this research. Lack of resources to regularly engage with participants across large geographical distances, including two participants who can only be reached through intermediaries, prevents this from approach being practicable. The need for such resources were not identified at the design stage. However, there is a commitment to revisiting the choice of pseudonym/actual name with each participant, as well as checking out the data analysis method with all, prior to submission of this doctoral thesis.
Methods of dissociating sensitive data from participants being practiced in this research include: not attributing names (authentic or not) to contentious quotes; summarising instead of quoting to remove colloquial terms; removing country references; and, when necessary, omitting particular data. A small number of comments were stated as being off the record, although this still leaves substantial data that was shared freely, but which may be harmful if associated with a particular participant. In full recognition that identifying data as sensitive, capable of causing harm, or compromising the integrity of a particular participant is a highly subjective process, the researcher is adopting a cautious approach. This is in response to the requirement of doing to harm, reducing the risk of which is so central to the ethics process. However, the author is aware that such caution may be at the expense of the participants’ presence, so important to activists such as Sarry and Rosa, and is therefore a compromise.