Overcoming ethical barriers to research

Abstract

Researchers engaged in studies about ‘hidden social groups’ are likely to face several ethical challenges. Using a study with undocumented Chinese migrants in the UK, challenges involved in obtaining approval by a university research ethics committee are explored. General guidance about how to resolve potential research ethics issues, with particular reference to ‘hidden social groups’, prior to submission to a research ethics committee is presented.

Introduction

Researchers engaged in examining social issues relating to ‘hidden groups’ face many ethical challenges. These challenges relate to hidden group vulnerability; difficulties in identifying and accessing ‘hidden participants’; and the possibility that research participation may stigmatize or re-traumatize participants. Researchers have addressed such issues head-on, used creative methods to overcome ethical constraints (see, for example, Cornwall and Jewkes’ 1995 review of participatory methodologies) or, regrettably, occasionally disregarded standard ethical principles (a notorious example is found in the covert methods used by Humphreys, 1975). In this article, we explore the challenges involved in obtaining ethical approval for research with one particular hidden group: undocumented Chinese migrants. First, we provide an overview of the project context, aims and methods. Next, we examine the nature of the issues raised by the University Research Ethics Committee (UREC). We argue that these issues are both technical and conceptual in nature and we outline the response taken to each type of issue. Finally, drawing on Guillemin and Gillam’s (2004) reflexive approach to research ethics, we analyse how general principles drawn from our experience about common ethical concerns can have relevance for qualitative researchers engaged with studies involving hidden groups.

The project

There is evidence both that the global population of irregular migrants has been increasing (Koser and Laczko, 2010) and also that undocumented status has been associated with significant psychological, social and political difficulties (Bloch et al., 2014). The issues addressed in this article arose from a UK study of 15 Chinese migrants with irregular migration status conducted by the first author and supervised by the second author. The study design comprised participatory interviews (conducted in English and Chinese), adapted from Wang and Burris’s (1997) photovoice model, in which participants used a provided digital camera to photograph images that conveyed their experiences of irregular migration status. These visual methods enabled participants to ‘construct accounts of their lives in their own terms’ (Holloway and Valentine, 2004: 8) and to prompt discussion of participants’ memories and feelings that interviews alone may not have evoked (Bagnoli, 2009; Harper, 2002).

The ethical issues

The project was submitted to the UREC, which identified three issues with both technical and conceptual components. These issues related to: the role of the researcher, the identification of participants and the nature of cross-language research.

Role of the researcher

Technical concerns

Committee members suggested that participants may be identified as criminals and queried whether the first author would have a legal duty to report disclosures of criminality. These concerns contained two key misunderstandings: first, that individuals who violate immigration laws are criminals when, in fact, the majority of immigration offences are civil rather than criminal offences (Aliverti, 2016); second, that researchers have an obligation to report disclosures of criminal offences made during research. ⁱ

In response, the revised second submission to the UREC clarified the legal obligations of researchers by reference to legislation and established UK ethical frameworks. For example, the authors cited the British Society of Criminology Statement of Ethics (2015) that states ‘In general in the UK people who witness crimes or hear about them before or afterwards are not legally obliged to report them to the police. Researchers are under no additional legal obligations’ (11). To add weight to this position, we drew the committee’s attention to previous empirical studies relating to irregular migration where researchers have protected the identities of participants with ‘illegal’ status (see, for example, Bloch et al., (2009) in the UK and Gonzales, (2011) in the US).

Conceptual concerns

The committee were concerned about how the first author’s professional identity, as a social worker, might affect the researcher–participant relationship. The UREC questioned whether the first author’s professional responsibilities to assist would become paramount if participants requested support or made disclosures of harm during or after fieldwork. According to Bell and Nutt (2012) based on their examination of practitioner research, differentiating the ‘researcher role’ from the ‘social worker role’ is difficult to achieve during fieldwork when unanticipated ethical dilemmas and emotions are likely to emerge.

In response, we provided the committee with a detailed account of how the first author would respond to safeguarding issues and participant requests for support during and after fieldwork (which involved signposting to health, legal and education support services). In addition, to enhance the first author’s ability to deal reflexively with emerging dilemmas, supervision with more experienced practitioner-researchers would be used to develop situated responses to fieldwork problems.

Identification of participants

Technical concerns

The UREC requested a detailed account of how participants’ identities would be protected, given that participation in the study could potentially reveal their migration status. The research design involved participant-produced visual data that would increase the visibility of participants who, out of necessity, actively hide their undocumented status. Therefore, the researchers were asked to clarify how photographs and other visual data would be anonymized and where they would be displayed (and for what purposes).

In response, the researchers drew on three arguments: recommendations of researchers in the field of irregular migration (such as Düvell et al., 2010); methodological debates about visual data (such as Wiles et al., 2008); and established ethical frameworks (such as the Statement of Ethical Practice for the British Sociological Association Visual Sociology Group, December 2006) to develop a protocol that would protect participants’ identities. This protocol comprised several practical arrangements in addition to the standard practice of removing real names, addresses and other identifiable details from data. For example, contrary to the accepted practice of obtaining signed consent forms, we would follow Düvell et al.’s (2011) recommendations by asking participants to indicate consent verbally rather than in writing. We would also arrange safe interview venues where participants would not be identified by others within their communities. Photographs produced by participants would be edited by the first author to remove identifiable people, objects and places and plans for the display of the photographs would be made clear to participants during the negotiation of consent.

Conceptual concerns

The committee members queried how the manipulation of participant-produced photographs for anonymization would affect the meaning that participants intended – a key dilemma for visual researchers. Empirical studies about researchers’ experiences of using visual methods have suggested that decisions about how to anonymize visual data (or not) are embedded in philosophical argument about researcher paternalism and participant autonomy (Wiles et al., 2012). The scientific validity of visual data that has been significantly edited and amended for anonymization also requires consideration. Although the researchers recognize both philosophical and scientific considerations, the vulnerability of participants due to their undocumented status had to be considered.

This conceptual question about anonymization prompted a significant amendment to the research design. Initially, the researchers planned to request consent for the use of participants’ photographs before anonymization. However, to provide participants with the opportunity to comment on their edited photographs, an additional consent process was developed that would take place two–four weeks after the participants’ final interview with the first author. Participants would have the opportunity to comment on their anonymized photographs and to consider whether they were satisfied (or not) that: a) their identities were sufficiently protected, and b) the integrity of their creative work had been maintained. Although this amendment to the research design did not resolve the committee’s concerns entirely, it increased participant control over the use of data and it served to reassure the UREC that the researchers were taking steps to manage the challenges involved in participant identification.

Cross-language research

Technical concerns

The committee members asked the researchers to provide clarity about the interpreter and translators’ qualifications, accountability and connection with the participants and their communities. Conducting research with interpreters and translators would complicate assurances given to participants about confidentiality by introducing additional actors into the researcher–participant relationship. If interpreters and translators were from the same community as participants, the risk of participants’ identities being revealed would be increased and this could affect the information that participants chose to share. The committee requested that the researchers develop an approach to ensuring that interpreters and translators adhere to the standards of confidentiality required in qualitative research with hidden social groups.

These concerns were addressed by a detailed account in the second submission to the UREC about interpreter and translator recruitment, working contracts and whistle-blowing procedures. An agreement of confidentiality for interpreters and translators was developed. If it transpired that participants and interpreters had a professional or personal relationship, interviews would be rearranged with an unknown interpreter.

Conceptual concerns

The committee also asked the researchers to consider how they would prevent important data from being ‘lost in translation’ given that translation is more than a simple technical process (Bassnett, 2013). According to Temple and Edwards (2002), their experience of conducting cross-language research revealed that language has cultural, political and social meanings, which create social realities that are difficult to translate directly between languages. Furthermore, the use of certain words, phrases or language forms have been found to be important in identity formation (Temple and Edwards, 2002). The interpersonal dynamics of the three-way interview between researcher, participant and interpreter can also, according to Edwards (1998), affect the data produced. Given the focus on examining participants’ qualitative experiences of undocumented migration status, the choices made during the processes of face-to-face interpretation and subsequent translation would affect how participants’ experiences were understood.

The researchers responded to these concerns in accordance with Edwards’ (1998) recommendation to increase the visibility of interpreters and translators in qualitative research. First, Edwards recommended working with interpreters before fieldwork to decide how interviews would be conducted: who would ask the interview questions and how would participants be encouraged to elaborate or clarify certain points. Second, Edwards recommended conceptualizing interpreters as ‘a form of key informant’ (1998: 203), which involves understanding the interpreters’ views of the research topic, their relationship with the researcher and participants and their thoughts about the interpersonal dynamics of interviews. These requirements would be satisfied by a pre-interview with the interpreter before fieldwork took place to understand their views of undocumented migration. Additionally, discussion between the first author and the interpreter following data collection would focus on linguistic judgements and the translation of culturally sensitive concepts.

These issues were further complicated by the first author’s status as an ‘outsider’ conducting research with participants from a minority ethnic population in the UK. The merits of ‘insider’ and outsider positions in qualitative research continue to be contested (see, for example, Hockey’s (1993: 199) review of the methodological debates on ‘going native’ and ‘going stranger’). Although the first author has acquired Chinese language and cultural knowledge through professional experience of working with Chinese families and two years spent living in mainland China, she would most likely be considered an outsider by participants given her white, British citizenship status and her position as an academic researcher.

To address concerns about the cultural competence embedded in the study design, several revisions were made. First, the researchers explained that the study was designed and developed in consultation with Chinese community workers to ensure that the focus of the project aligned with the needs of the UK Chinese population. Second, during the data analysis process, the researcher would receive supervision from Chinese social-work academics from the same language and cultural group as the participants. The involvement of Chinese community advisors and Chinese social-work academics was designed to challenge the cultural bias of the researchers.

Discussion

It is important to acknowledge that all ‘the day-to-day ethical issues that arise in the doing of research’ (Guillemin & Gillam, 2004: 264) cannot be predicted in advance. However, in our experience, the ethical uncertainty implicit in research with hidden groups can be managed within a methodological design that creates opportunities for researchers to discuss issues as they emerge with experienced colleagues and advisors from the same cultural and linguistic group as the participants. We take it as axiomatic that researchers (proposers) who submit a research proposal focussed on hidden groups to an UREC, of whatever type, will want their proposal to be approved ideally on the first submission, but certainly with as few requirements for amendment as possible. Two general guidance notes can be drawn from our experience of overcoming ethical barriers to research:

1. At the outset, and prior to the submission to the UREC, the proposers should attempt to distinguish those aspects of the proposal that are technical in nature from those that are conceptual. Ethical issues that are technical in nature can be satisfied with full and detailed preparation of the case, taking account of potential technical, legal and practical issues (e.g. technical processes involved in the management and anonymization of data, and the arrangements in place to manage working relationships with other actors, such as interpreters or gatekeepers). Proposers should avoid making assumptions about UREC members’ knowledge about the legal obligations of the researchers in studies with hidden social groups; rather these obligations should be explained explicitly with reference to relevant legislation, disciplinary codes of ethics and the approaches used by established researchers in the field.

2. Where, prior to submission, conceptual issues can be identified in the proposal (e.g. relational aspects of fieldwork, professional sensitivities of researchers, and processes involved in interpretation and representation of participant views), these may be comparatively more difficult to resolve in advance than the technical issues. One method of addressing these issues, drawing on Guillemin and Gillam’s (2004) recommendation, is to adopt a reflexive approach to research ethics. During the development of research ethics proposals for studies with hidden groups, researchers can develop ethically sound and reflexive practices by questioning their own role (and the role of others), the aim of the research and the impact on participants (Guillemin and Gillam, 2004) and reflect these in the UREC proposal.

Conclusion

Our experience has demonstrated that overcoming ethical barriers to research with hidden groups, raised by submission to the University UREC, involved responding to both technical and conceptual issues relating to research ethics. We acknowledge that, by definition, our approach to addressing ethics in practice issues is partial, as it is impossible to anticipate all eventualities before beginning fieldwork. However, we have argued that qualitative researchers can overcome barriers to research with hidden groups by developing technical accounts of their studies that include descriptions of the legal duties of researchers, and by developing reflexive approaches to address the ethical issues that are likely to emerge in practice. Finally, we have suggested that research design should create space for dialogue between researchers and the wider communities of the hidden group to address the unanticipated ethical issues that emerge during fieldwork and to challenge researchers’ cultural bias.