A Clash of Paradigms? Ethnography and Ethics Approval

Abstract

Obtaining ethics approval from university ethics committees is an important part of the research process in Australia and internationally. However, for researchers engaging in ethnographic work, obtaining ethics approval can (re)present significant hurdles to overcome in planning and facilitating a research project. In this article, we discuss potential challenges of reconciling the differences between institutional ethical review standards and the reality of ethnographic research. To do so, we reflect on our own experiences seeking ethics approval for a study on racialized visibility in rural nursing and another on the experiences of gender and sexuality diverse older women. We focus on two particular queries from ethics committees that reaffirm, for us, the incompatibility of biomedically informed ethics guidelines for naturalistic, ethnographic research. The article draws on four major points of contention regarding ethical approval processes designed for biomedical research and applied to social research. With respect to social research, these are (a) the associated risks, (b) predictive informed consent, (c) the power held by social researchers, and (d) biomedical emphasis on distance and universalism within the research relationship. This article suggests a reformulation of ethics guidelines and structures such that ethics committees are better able to engage with ethnographic (and other social) research. Although these debates and structural changes may not be relevant for all social or ethnographic research, exploring these ethical difficulties is paramount to redefining expectations and the positivist standards upon which social research is often measured.

Introduction

[E]thical guidelines and the institutional review system are prejudiced in favor of research methodologies that mimic epidemiology, and kindred “objective” approaches that deal with humans as detached social atoms. These approaches do not address the needs or practices of participant observation.

—Oeye, Bjelland, and Skorpen (2007, p. 2296)

Obtaining ethics approval from university Human Research Ethics Committees (HRECs) and/or institutional review boards (IRBs) is an important part of the research process in Australia and internationally. However, for researchers engaging in ethnographic work, obtaining ethics approval can (re)present “significant hurdles to overcome in planning and facilitating a research project” (Davison, Brown, & Moffitt, 2008, p. 2). This may be especially true when that research occurs within health care settings, where issues of (patient) vulnerability, as a result of poor health status and practitioner–client relationships, abound. As Librett and Perrone (2010) point out, “[w]hile Institutional Review Boards (IRBs) and ethnographic research seek to protect the participants of research, they operate on two diametrically opposed paths” (p. 279).

Not surprisingly, therefore, a small but expanding body of work into this seemingly contradictory relationship between the “heavy procedural requirements” (Davison et al., 2008, p. 2) of HRECs and the nature of ethnography now exists. In fact, the background literature review for this article identified a peer-reviewed journal special issue dedicated to the subject of informed consent and the processes by which this is reviewed and obtained (Social Science and Medicine, Vol. 65, No. 11). Notably, this special issue discusses a range of topics related to the transformation of ethical and informed consent procedures including the meaning, role, and requirements for consent as well as the biomedical paradigm of ethics procedures and their impact on social science research. With regard to the latter, the special edition editors argue that

the values, concerns and traditions of biomedicine have come to dominate research ethics committees, wherever they are located, in terms of their membership, operating procedures and the framework within which they review proposals. And it is this dominance of the biomedical paradigm, and the poor understanding of social science research among members of research ethics committees, that have tended to be the focus of criticism among social scientists. (Boulton & Parker, 2007, p. 2190)

This position is highlighted by the articles included within the special issue which present four main arguments against the application of a biomedical research approach to ethics in social science research.

First, social scientists argue that social research is associated with a lower order of risks. While social research may be intrusive, it is not invasive and does not generally involve the risk of direct physical harm, injury, or death (Boulton & Parker, 2007). Furthermore, it is argued that the risks of social research are similar to those that humans encounter daily and from which we do not require the same level of protections as we might in medical research (Boulton & Parker, 2007). As such, social scientists may perceive the formal review and consent procedures required by ethics committee to be heavy handed and/or excessive (Whittaker, 2005; Williamson, Kent, Goodenough, & Ashcroft, 2002).

Second, the requirement for “anticipatory” or “predictive” informed consent as is usual in biomedical research is not always possible or even desirable within social science research (Strathern, 2000). As explained by Miller and Bell (2002), for qualitative research, especially those using ethnographic methods, neither the researcher nor participants can anticipate how the research will develop or what issues may arise, making it near impossible to specify all the risks involved in participation in advance of the study.

Third, unlike the biomedical research context, social researchers often hold less social capital (power) than their biomedical researcher counterparts which allows for “more enduring, negotiated and equal [relationships], affording participants a greater degree of agency and requiring researchers to reflect continuously on ethical dimensions of their activities” (Boulton & Parker, 2007, p. 2191). Contrastingly, Boulton and Parker (2007) argue that participants’ willingness to take part in medical research may be influenced by the institutional context and affiliation of researchers while social science researchers must establish their credibility and trustworthiness on their own merit. As Springwood and King (2001) explain, “‘getting along’ with others anchors the ethnographic process, precisely because the significance of other worlds materializes through others’ words” (p. 407). Given the evolutionary nature of the relationship between researcher and participants, many social scientists see consent as an open-ended process that is constantly negotiated (Ramcharan, 2006). Without this fluidity and flexibility, the trust required for robust data to emerge is disrupted. In these contexts, “trust is earned incrementally . . . [and] . . . the requirement for signed consent forms, which constructs the relationship in a legalistic way, is inappropriate . . .” (Boulton & Parker, 2007, p. 2191).

Finally, many researchers, including those interested in bioethics, are challenging the current rights-and-justice-based model of ethics that underpins informed consent. Such researchers object to the emphasis of distance and universalism as the basis for the research relationship and instead suggest an ethical relationship based on particularism, collective rights, and active engagement (Edwards & Mauthner, 2002). In this regard, it is argued that instead of focusing on specifying and eradicating all risks related to the study, a priori attention should be moved to engaging actively with ethical dilemmas as they arise—moving away from individualistic approaches to those that are more relational (Boulton & Parker, 2007). These arguments and procedural changes may not be relevant for all social research or in fact all ethnographic studies. However, exploring the “ethical difficulties when trying to establish informed consent among research participants” (Davison et al., 2008, p. 2) is integral to redefining ethics committee expectations and the biomedical yardstick upon which social research is measured.

In line with the aforementioned areas of debate, this article examines the relevance of ethics approval processes (guidelines and requirements) and eventually, obtaining ethics approval, for ethnographic studies conducted in health care settings—especially within an Australian context. In doing so, this article adds to a small but emerging body of work (Dickson-Swift, James, Kippen, & Liamputtong, 2008; Dodds, 2002; Komesaroff, 2002; Melville, 2005; Roberts, Henderson, Willis, & Muir-Cochrane, 2013).

The article is informed by our experiences trying to obtain ethics approval for projects using ethnographic methods to explore the experiences of marginalized populations across a range of health contexts including nursing, medicine, and allied health in Australia. Specifically, the article uses as the reference point for debate our discomfort with two “queries” raised by HRECs in relation to some of our projects: (a) that of prior informed consent, not only of the participants of our research but all individuals who may be “accidentally” observed by virtue of being in close proximity with the participant and (b) the expectation that ethnographic research practice must be objective, and have clear directions and measureable outcomes.

The Studies: Ethnographic Method

The ethnographic projects on which this article is based utilized participant observation of non-White nurses and gender and sexuality diverse (GSD) older women. In addition to methodological similarities, both studies focused on the experiences of minority and under-researched populations who experience social, economic, political, and cultural marginalization manifesting in prejudice and discrimination. With respect to research, these experiences and/or social positioning may make them sensitive population groups—a central concern for ethics committees. Each study along with its aim and method are described below.

Study 1

This project occurred in nursing units in two rural hospitals as examples of multicultural workplaces constituted in part by the societal beliefs about race, ideologies of multicultural tolerance, and a raft of workplace policies in support of equality and diversity management. The aim of the study was to critically analyze the social relations of black nurses within these workplaces—claimed as multicultural and cosmopolitan—to begin developing nuanced understanding(s) of how racial visibility is constructed as problematic difference. Our decision to use organizational ethnography was predicated on our belief that the unarticulated meanings of difference constructed within broader workplace policies are sometimes enacted within normalized acts of microaggression based on racial visibility. The study used participant observation of nurses who identified as racially visible as they went about their day-to-day work. A review of diversity management policies as well as nurses’ interactions with colleagues, patients, and their relatives was analyzed for what they exemplified as to how the worker, the nursing unit space, and the organizational rules becomes racially coded.

Study 2

This project engaged with GSD (e.g., lesbian, bisexual, transgender, intersex, or queer) women 55 years and older and explored their experiences with health, social, and aged-care services. Given that these services, including local health centers, community social support, community mental health support, and home care, are designed to safeguard clients’ well-being and social connection, it was imperative to understand how interactions with these services may either support or work in direct contrast to the service’s mission. Rapid ethnographies (Baines & Cunningham, 2013) with the women were conducted to clarify the needs of these women and to design a quantitative survey tool for data collection and qualitative questions for an online forum where the women could interact with one another. The questionnaire enabled participants to report on inclusive practices of the health and social services that they have accessed. The online forum discussions allowed for a better understanding of participants’ self-directed help-seeking behaviors and how these practices both complement and supplement face-to-face interactions with service providers.

Query 1: Informed Consent

The bottleneck for social scientists and especially ethnographers seems to be the issues surrounding informed consent. Participants are required to be notified, in writing, of the potential risks and benefits of their cooperation (Librett & Perrone, 2010).

As indicated, informed consent is predicated on the argument that potential participants need to not only consent to participate in the study, but to know what they are agreeing to, that is, be fully aware of what it means to participate in the study, and have complete knowledge of what the study is about.

Moreover, ethics committees require not only that participants make informed decisions on whether or not to participate in a study, but that this consent is written and acquired before participation in any aspect of the research. This rigidity is especially remarkable considering that Section 3.1.16 of the National Statement on Ethical Conduct in Human Research published by Australia’s National Health and Medical Research Council (2007) stipulates that

[t]he method of providing consent in qualitative research depends on various factors, including the type of research, its level of sensitivity, its cultural context, and the potential vulnerability of the participants. In some contexts, the protection of vulnerable participants may favour a formal, written process of consent; in other contexts, an oral process. (p. 28)

Like many ethnographers (see, for example, Corrigan, 2003; Murphy & Dingwall, 2007; Roberts et al., 2013), we find this rigid, biomedically informed notion of “informed consent” limiting and potentially impinging on our study process as we discuss below.

Our Case

A significant impetus for this article, as mentioned above, was consternation with the requirement to obtain prior informed consent from all those in contact with the nurses and GSD women who may be observed, before the studies commenced. According to the ethics committee, prior consent of all colleagues was necessary in case other individuals were (accidentally) observed in the ethnographic process. Despite this consternation, a response that met the ethics committee’s requirement had to be submitted for approval to be granted. In the case of the GSD women, observation during participants’ daily activities, including appointments at social, aged care, and health services, was abandoned. As per the ethics committee, their inclusion would require additional ethical approval from external organizations and therefore logistically implausible. In the case of the nurses study, two changes had to be made to the protocol. First, an “additional non-participant group” was added in response to the question of how many “groups of participants” would be involved in the study. This group comprised of colleagues who were not participants per se (i.e., who would neither be interviewed nor be the focus of observations, whose inclusion in the study was only incidental, due their proximity to those who were being observed). Second, a “non-participant” group consent form had to be designed and submitted to the ethics committee.

The problem

Our main problem with the ethics committee’s requirement is not with obtaining consent from individuals who may be “accidentally” observed. We do not make light of or object to obtaining such consent. Rather, our concern is that this consent must be prior consent, and obtained from all individuals and organizations, which for us means obtaining consent from people who may never be observed. This requirement by the ethics committee, we argue, highlights not only a clear lack of understanding that in ethnographic work consent is an ongoing negotiation that occurs throughout the research process, it also exposes the application of “a tightly controlled positivist lens” (Librett & Perrone, 2010, p. 731) to what “consent” means in research.

As Librett and Perrone (2010) observe, “[a]dherence to the rote insistence of many review boards to standards of informed consent more appropriate to medical experimentation is a pressing issue for ethnographers” (p. 733). Justifiably, therefore, our concern is that any attempt to obtain consent from all individuals and organizations before we commenced the studies and before we established rapport with our participants and their colleagues was likely to fail. As previously noted, the issue of rapport between the researcher and participants is central to the ethnographic endeavour, particularly when seeking access and obtaining consent. Glesne (1989) concurs:

Rapport’s contribution to all modes of ethnographic research is essential. The researcher could not succeed without the trust that rapport engenders. We ask others to allow us to watch and record what they do and say. We ask them to describe and explain what they may never before have given voice to. (p. 53)

To us, therefore, the ethics committee’s requirement for prior informed consent clearly shows a lack of understanding of the process and social relationships of fieldwork, the instrumental role of rapport in these processes and relationships, as well as its “importance . . . for the purpose of ‘producing’ knowledge” (Springwood & King, 2001, p. 407). However, the requirement of a priori consent ignores the intimacy involved in our relationships with participants and assumes that all individuals will be observed and engaged to the same extent as our participants. Previous work in a burns unit by the lead author highlights that not all colleagues are observed (Mapedzahama, Rudge, West, & Perron, 2011). If that is the case, our argument is that prior consent from all individuals is a futile exercise.

We, therefore, question how ethical it is to obtain consent from people who may not be involved in the study. Moreover, if we cannot obtain prior consent from someone who may not even be “accidentally” observed, does this mean we cannot commence observations of the participant? For us, this requirement not only assumes that all individuals will be involved in the study, albeit as the “accidentally observed,” but that this may expose them to some risk, no matter how small. This suggests that ethics committees can “potentially function as a ‘virtual police force’ running amok, and claiming jurisdiction over any possible contingency—however unlikely” (Librett & Perrone, 2010, p. 734).

As such, ethics committees “often presume[s] that subjects are the most at risk” (Librett & Perrone, 2010, p. 735), yet oftentimes, the ethnographer is exposed to greater risk than the participant, for example, in studies where they “bear witness to extreme risk-taking behaviours” (p. 739). While our studies do not inherently fall into this category, we argue, nevertheless, that our observations expose our participants to no greater risk than those observations expose us.

Finally, we query the meaning of informed, in “informed consent,” as understood by ethics committees. Murphy and Dingwall (2007, p. 2226) point out,

Since Nuremberg, there has been general agreement between ethicists, regulators and scientists that research participation must not only be voluntary but also based on adequate knowledge and understanding of the nature, duration, purpose, methods and potential hazards of the study [emphasis added].

Given that ethnography is an “iterative process where, typically, the researcher sets out with a broad question to be investigated and the precise lines of analysis emerge as the data accumulate” (Murphy & Dingwall, 2007, p. 2227), we question then, our ability to fully or adequately “inform” our participants. How informed can (or should) our participants be about research whose methodology (ethnography) intersects with silenced subject matter in Australia (race/racism, LGBTIQ identity, or aging/ageism)? In the end, this requirement for us represents the tendency for ethics committees to view the (process of) obtaining informed consent as a “clean,” sanitized, and straightforward process.

Query 2: Objectivity, Clear Directions, and Measureable Outcomes

In positivistic research structured observation is a discrete activity whose purpose is to record physical and verbal behaviour. Observation schedules are predetermined using taxonomies developed from known theory. (Mulhall, 2003, p. 306)

Our Case

In line with social researchers who debate the imposition of positivistic frameworks onto social research, we found further contention with the ethics committee’s queries regarding the deliverables and measurable outcomes being produced from our research. The ethics committee reinforced this query as they felt that the observations lacked direction and seemed to be a “series of rambling events and interactions.” This was supported by the requirement to specify to the assessing ethics committee, the exact time that would be spent conducting observations. Furthermore, the presence of the researcher within the ethnographic settings was questioned as the ethics committee was concerned that this would skew the observational data.

Again, to comply with ethics committee’s requirements, an amendment was made to the research protocols and Participant Information Sheet documents. In the case of the ageing GSD women, we needed to explain to the ethics committee that qualitative research does not problematize the presence of the researcher or their engagement, nor aim to extract the researcher, from participant interactions (Chenail, 2011). Responses to the ethics committee, therefore, required further explanation of ethnography and qualitative research as scientific methodologies in which the researchers’ positionality is declared and that an intimate participant–researcher relationship is ideal for the extraction of rich and multilayered data (Chenail, 2011). For instance, we paraphrased Hammersley and Atkinson (2007), who note,

The ethnographer participates, overtly or covertly, in people’s daily lives for an extended period of time, watching what happens, listening to what is said, asking questions; in fact collecting whatever data are available to throw light on the issues with which he or she is concerned. (p. 2)

The ethics committee was informed that these types of data could not be accessed if the researchers disconnected themselves from participants to maintain scientific objectivity (Weber, 2015). The ethics committee was also given more information about how ethnographic research evolves and that the directions of the research were to be decided by the participants who effectively guide the researchers through a day in their lives. To address the issues of deliverable and measurable outcomes, an observation mapping guide was included to facilitate documentation of the rapid ethnographies in a nationally standardized format.

With respect to observation times and in the case of the non-White nurses, both the research protocol and the Participant Information Sheet were changed such that it stipulated an average of 4 hr for each observation three times (i.e., in three shifts) each fortnight. Similarly, the GSD women study documents needed to indicate that the rapid ethnographies would take 4 to 6 hr during a day and location(s) of the participants’ choosing. In both studies, the researchers were guided by the participants and resulted in observations that ended up taking longer than initially indicated. Although these stipulations satisfied the ethics committee’s requirements and clearance was granted, they were especially unsettling for the authors for several reasons.

The problem

In our perspective, the ethics committee’s queries and requirements showed a lack of understanding of ethnographic research and qualitative methodologies. Notably, an acknowledgement of the different power relationships in ethnographic work, compared with other types of methodologies, was overlooked. As mentioned, biomedical research relationships with participants may be hierarchal with the researcher holding more power over the study’s directions, timelines, deliverables, and outcomes (Murphy & Dingwall, 2007).

However, in social research, the distribution of power is not always tilted in favor of the researcher; oftentimes, the participants command greater power over the research process: its progression as well as the amount and quality of data the ethnographer gets access to. Perhaps, most importantly, these requirements highlight a clear lack of understanding of what it means to “observe” ethnographically—in our research, a process led by the participants. In fact, Murphy and Dingwall (2007) suggest the metaphor of “hosts” and “guests” to describe the research relationship and the power dynamics in ethnography:

We prefer the terminology of “hosts” here because it better captures the nature of the relationship: ethnographers are guests in someone else’s setting and, like guests every-where, there are clear, if not always explicitly articulated, expectations of proper behaviour. The status of guests is always fragile and depends on appropriate conduct towards their hosts, although this is not necessarily reciprocated. (p. 2225)

As such, our discomfort with the ethics committee’s requirements stems from our contention that delineating all aspects of the project before it begins and measuring data and outcomes using positivistic frameworks assume that all participants experience their environments in the same ways and that those environments are static. Social environments such as nursing units, for instance, are versatile, fluid spaces whose constitution is ever changing. In the case of our non-White nurses, staffs from other units and also agency nurses are constantly coming in as and when required to boost numbers (meet nurse–patient ratios and the needs of patients). In acknowledgement of this fluidity, the ethics committee may have been concerned about the researcher polluting the data from a positivist perspective. However, naturalism is more in line with ethnographic methods as Hammersley and Atkinson (2007) explain:

Naturalism states that as far as possible, the social world should be studied in its “natural” state, undisturbed by the researcher. Hence, “natural,” not “artificial” settings like experiments or formal interviews, should be the primary source of data. (p. 6)

Given the nature of ethnographic research and the fluidity of the “natural” social state, which are participants’ lives, it is impossible to indicate a priori the exact directions in which the observations will go, what deliverables will eventuate, precise outcome measures, or how long observations will take. In regard to this latter point, the times specified were pure guesses. Given that ethnography emerges out of negotiations and re-negotiations in the field, it is virtually impossible to predict how much time one would spend in the field before the fact. For instance, with the non-White nurses study, this is an issue the ethics committee seemed to miss, even after a lengthy conversation with one representative, explaining the nature of ethnographic work, and the author was “encouraged” to “just put something down even if it is really broad.”

As such, specifying these aspects gives participants, researchers, and ethics committees an unrealistic point of reference for what participating would entail as well as what the research can and will produce. As with other social researchers, we purport that ethnographic research and the evolution of the aforementioned aspects are decided through negotiations with the participants. For us, making these decisions on behalf of participants places a restriction on the ethnographic process and potentially, the quality and amount of data gathered.

Conclusion

[T]he process [of ethics approval] has become so highly bureaucratized that, intentional or otherwise, the effect upon potential research directions has become undeniable, and there is a dire need to address the implications for academic freedom and the production of knowledge. (Librett & Perrone, 2010, pp. 729-730)

In this article, we have discussed the challenges of “reconcile[ing] the differences between institutional ethical review standards and the reality [of ethnographic research]” (Davison et al., 2008, p. 1). Drawing from our own experiences, seeking ethics approval for a study on racialized visibility in rural nursing and the experiences of GSD older women in Australia, we have discussed two particular requirements by ethics committees for our ethnographic studies that reaffirm for us the incompatibility of biomedically informed ethics guidelines for naturalistic, ethnographic research.

The article draws on previous research with particular focus on a special issue within the journal of Social Science and Medicine on the topic of ethics and social science. Included articles present four major points of contention regarding ethical approval processes designed for biomedical research and applied to in social research. These debates purport that the following:

1. Social research is associated with a lower order of risks. As such, social scientists may perceive formal review and consent procedures ethics committee to be heavy handed and/or excessive (Whittaker, 2005; Williamson et al., 2002).

2. Predictive informed consent is not always possible or even desirable within social science research (Strathern, 2000).

3. Social researchers hold less power than biomedical researcher counterparts which allows for “more enduring, negotiated and equal [relationships], affording participants a greater degree of agency . . .” (Boulton & Parker, 2007, p. 2,191).

4. The positivist and biomedical emphasis of distance and universalism as the basis for the research relationship dilutes the quality and quantity of data social research can obtain. Furthermore, an ethical relationship can be achieved with participants without having to specify aspects of all deliverables and outcomes a priori (Edwards & Mauthner, 2002).

Using these points to debate two sets of queries regarding our ethnographic research, this article has highlighted how poor understanding of ethnography by ethics committees/members and biomedical bias of the ethics guidelines have the potential to not only alter the “study methodology and direction of research” (Librett & Perrone, 2010, p. 736), but to do so in ways that may affect the quality of data collected, thus limiting the trustworthiness and impact of the research.

In other international settings (e.g., some universities in the United Kingdom, Canada, and across Europe), however, such incompatibility or incongruence may not be present (or significantly reduced) in the context of ethics committees that recognize the fluidity of ethnographic (and qualitative research) and had integrated this fluidity into their review processes (Pollock, 2012). This article, therefore, suggests a reformulation of ethics guidelines and ethics committee structures that reflect a range of research paradigms. In doing so, ethics committees may be better able to engage with the ethnographic (and other social) research in ways that decrease their need for extensive explanations of methodological and theoretical frameworks. In line with Librett and Perrone (2010),

We suggest improving and enhancing communication between review boards and ethnographers, and/or creating an alternative review process for those research endeavors (such as ethnographies), which lie well beyond the positivistic box. (p. 730)

Further engagement with existing resources for such a restructure, especially in the Australian context, can be developed using international guidelines. For instance, The Association of Social Anthropologists (1999) Ethical Guidelines for Good Research Practice outlines ways for thinking about informed consent as an ongoing process. These guidelines reiterate the importance of trust and rapport within the context of social research and highlights means for managing the fluid nature of some social research—a benefit for ethical review processes and managing ethics committee concerns. A full description and discussion of these can be found at https://www.theasa.org/downloads/ethics/Ethical_guidelines.pdf

Such suggestions for restructuring may therefore reduce the need for ethnographic and social researchers to make major changes to or abandon research that does not fit neatly into a positivist framework. Although these debates and structural changes may not be relevant for all social or ethnographic research, exploring these ethical difficulties is paramount to redefining expectations and the positivist standards upon which social research is measured.