ETHICAL -ADHA

10.1177/1077800403262360 ARTICLEQUALITATIVE INQUIRY / April 2004Guillemin, Gillam / “ETHICALLY IMPORTANT MOMENTS”

Ethics, Reflexivity, and “Ethically Important Moments” in Research

Marilys Guillemin Lynn Gillam

University of Melbourne

Ethical tensions are part of the everyday practice of doing research—all kinds of research. How do researchers deal with ethical problems that arise in the practice of their research, and are there conceptual frameworks that they can draw on to assist them? This article examines the relationship between reflexivity and research ethics. It focuses on what con- stitutes ethical research practice in qualitative research and how researchers achieve ethi- cal research practice. As a framework for thinking through these issues, the authors dis- tinguish two different dimensions of ethics in research, which they term procedural ethics and “ethics in practice.” The relationship between them and the impact that each has on the actual doing of research are examined. The article then draws on the notion of reflexivity as a helpful way of understanding both the nature of ethics in qualitative research and how ethical practice in research can be achieved.

Keywords: research ethics; reflexivity; research practice

Picture this scene. You are a researcher working on a study examining women’s experiences of heart disease. You are interviewing Sonia, a woman in her late 40s with diagnosed heart disease. Sonia lives on a remote farming property in a rural region. She is married and has one teenage daughter living at home with herself and her hus- band. The interview is progressing well. Over a cup of tea in Sonia’s kitchen, you inquire about the impact of heart disease on her life. Sonia stops and closes her eyes. After a few moments’ silence, you notice tears welling up in Sonia’s eyes. Sonia tells you that she is not coping—not because of her heart disease, but because she has just found out that her husband has been sexually abusing her daughter since she was a child.

This kind of scenario is not unusual when conducting qualitative re- search. Most qualitative researchers can describe similar experiences they

261

Authors’ Note: We would like to gratefully acknowledge the suggestions and com- ments of our colleagues at the Reflexivity seminar series at which a draft version of this article was presented.

Qualitative Inquiry, Volume 10 Number 2, 2004 261-280 DOI: 10.1177/1077800403262360 © 2004 Sage Publications

have encountered, usually with considerable emotion and crystal-clear re- call. The issues raised in this scenario have ethical and legal implications. How as researchers do we respond to such disclosures when they occur, and are there existing conceptual frameworks and principles that we can draw on to assist us? Our focus in this article is on what we refer to as the “ethically important moments” in doing research—the difficult, often subtle, and usu- ally unpredictable situations that arise in the practice of doing research. We are interested in the ethical practice of research and how this is achieved. We examine existing ethical principles and frameworks for both their limitations and what they offer researchers and then turn to reflexivity as a potential tool for ethical research practice.

Ethical dilemmas and concerns are part of the everyday practice of doing research—all kinds of research. Ethics is certainly not confined to qualitative research or necessarily to research that involves humans. Ethical issues are equally pertinent in clinical trials that are primarily quantitative, in research that involves animals rather than humans, in research that involves humans only indirectly (such as that which investigates human remains or documents that relate to people), and even in nonempirical research where the indirect or long-term consequences can be significant. The traditional starting point and focus for discussions of research ethics has been the ethical principles for- mulated for biomedical research, usually quantitative in nature. However, since the 1960s, social scientists have also explored ethics in qualitative social research (Bulmer, 1982; Homan, 1991; Kimmel, 1988; Orb, Eisenhauer, & Wynaden, 2001; Robley, 1995). In this article, we build on this work by focus- ing on what constitutes ethical research practice in qualitative research and how researchers achieve ethical research practice. Specifically, we examine the particular concerns of ethics in qualitative research and how they relate to biomedical research ethics as traditionally defined. In particular, we ask whether there are theoretical resources within the tradition of qualitative research for dealing with ethical aspects of research practice or whether these resources must be drawn from elsewhere. This question arises in a context where externally generated principles of research ethics have been applied to qualitative social research and where disquiet about the appropriateness of this has periodically been expressed (American Association of University Professors, 2000; Bouma & Diemer, 1996).

As a framework for thinking through these issues, we distinguish two dif- ferent dimensions of ethics in research, which we term procedural ethics and “ethics in practice.” We examine the relationship between them and the impact that each has on the actual doing of research. Having shown that pro- cedural ethics cannot in itself provide all that is needed for dealing with ethi- cally important moments in qualitative research, we then draw on the notion of reflexivity as a resource. This is the key step. Although reflexivity is a famil- iar concept in the qualitative tradition, we suggest that it has not previously been seen as an ethical notion. We propose that reflexivity is a helpful concep-

262 QUALITATIVE INQUIRY / April 2004

tual tool for understanding both the nature of ethics in qualitative research and how ethical practice in research can be achieved.

DIMENSIONS OF ETHICS IN RESEARCH

We suggest that there are at least two major dimensions of ethics in qualita- tive research (indeed in all research, but we will not pursue that issue here). These are (a) procedural ethics, which usually involves seeking approval from a relevant ethics committee to undertake research involving humans; and (b) “ethics in practice” or the everyday ethical issues that arise in the doing of research. It could be argued that there is another dimension— research ethics as articulated in professional codes of ethics or conduct. Most professions and organizations have professional codes of conduct (Bulmer, 1982; Coady & Bloch, 1996; Homan, 1991). Although these codes do exist and are often essential components of the constitution of professions and orga- nizations, we, like other analysts such as Mason (1996, p. 166), would ques- tion the relevance of these codes for actual research practice. Hornsby-Smith (1993) claimed that organizational codes of conduct are often too restrictive, and that there is a general move in many organizations toward more general guidelines for ethical practice. In terms of usefulness in addressing ethical issues that arise in practice, professional ethical codes are largely not practical or applicable and can serve only as general guidelines. Although we do not wish to totally dismiss the role of professional codes in guiding ethical research practice, we focus in this article on procedural ethics and “ethics in practice” and the relationship between them. We begin by reflecting on the impact that these different dimensions actually have on research.

Procedural Ethics

In qualitative research that involves humans, one of the early stages of the research process is the completion of the application form for a research ethics committee.1 For many researchers, the completion of the research ethics com- mittee’s application form is a formality, a hurdle to surmount to get on and do the research. Like many of our qualitative research colleagues, we diligently answer the questions on the ethics application form, even though they may be irrelevant to our research. We have learned to write our responses to the ques- tions in “ethics-committee speak.” This involves using language that the committee will understand, is free of jargon, but will nonetheless reassure the committee that we are competent and experienced researchers who can be trusted. This also involves explaining methodology to a committee who may be unfamiliar with qualitative methods and in some cases, may be antagonis- tic toward this type of research. Moreover, we have learned to gloss over some

Guillemin, Gillam / “ETHICALLY IMPORTANT MOMENTS” 263

issues that we know may cause the committee concern, for example, giving a transcriber, who is external to the research team, access to interview tapes without seeking direct consent from the participants, or not to draw too much attention to potential, although unlikely, risks to the researchers when con- ducting the research (in particular, risks to research assistants or less- experienced researchers on the research team). The form asks what measures the researcher/s have put in place in the event of unexpected outcomes or adverse effects. As indicated in the opening scenario, there are many situa- tions that are unexpected when doing research that can potentially have adverse consequences—how can you foresee and plan for all of these? Most researchers learn quickly that they need to be savvy in addressing the poten- tial issues of concern of the committee: using the appropriate discourse to en- sure that applications will be approved as quickly as possible with minimum changes and dispute, while remaining true to their research integrity.

“Ethics in Practice”

The second dimension of ethics we consider is “ethics in practice.” These pertain to the day-to-day ethical issues that arise in the doing of research. Consider the example of Sonia’s disclosure that her husband has been sexu- ally abusing her daughter. This can be interpreted as a classic ethical dilemma of whether to breach Sonia’s confidentiality to prevent harm to her daughter. Even with this interpretation, this is not an ethical question that receives much attention at the procedural ethics level, where the focus is much more on how confidentiality will be maintained rather than on when a breach of confidentiality might be ethically required. More important, we suggest that this is not the only ethical issue here or even the most pressing one. There are many more immediate ethical concerns. The researcher has to decide how to respond to what Sonia has said. Does the researcher let the disclosure pass or take it up in some way? And in what way—what words to say, what tone of voice to use? Turn off the tape recorder or keep it running? Abandon the inter- view plan or try to return to it? Offer to discuss the situation or offer to help in some way? These are issues about the ethical obligations a researcher has toward a research participant in terms of interacting with him or her in a humane, nonexploitative way while at the same time being mindful of one’s role as a researcher.

These issues are not usually addressed in research ethics committee appli- cations, nor are they events that are often anticipated when applying for approval. Some researchers may not even regard them as ethical in the sense that they may not be “dilemmas” or appear to be of great consequence. It is certainly true that some of them are not dilemmas, if we take dilemma to refer to a situation in which there is a stark choice between different options, each of which seem to have equally compelling ethical advantages and disadvan-

264 QUALITATIVE INQUIRY / April 2004

tages. Perhaps these issues may seem unimportant because they have an “everyday” sort of quality. However, there is much more to ethics than red- letter dilemmas, and much is missed by not being aware of this. The questions that we have posed about responding to Sonia represent what we call “ethi- cally important moments,” where the approach taken or the decision made has important ethical ramifications, but where the researcher does not neces- sarily feel himself or herself to be on the horns of a dilemma. In fact, in some cases, it might be relatively clear how the researcher should respond or pro- ceed, and yet there is still something ethically important at stake. For exam- ple, it might be clear that the researcher should not just continue with the interview as if nothing has happened but instead, should respond directly to Sonia in some way. That this is clearly the right thing to do does not make the situation ethically trivial; the moment of response is an ethically important moment for there is the possibility that a wrong could be done. There can be all sorts of ethically important moments: when participants indicate discom- fort with their answer, or reveal a vulnerability; when a research participant states that he or she does not want to be assigned a pseudonym in the writing up of the research but wants to have his or her real name reported; or the case described by Orb et al. (2001) of interviewing victims of violence where the researcher has to decide how far to probe a participant about a difficult and distressing experience.

It is this dimension of ethics made up of “ethically important moments” that is of primary interest in this article. Although this ethical dimension of research practice is often apparent to researchers, there is little conceptual work available to draw on to make sense of it. We need both a language to articulate and understand these ethical issues and an approach that assists us to deal with these issues when they arise. We will now go on to suggest a way into this “ethics in practice” dimension and then consider how “ethics in practice” relates to procedural ethics.

Microethics in Research Practice

A potentially useful term for thinking about “ethics in practice” is micro- ethics. This term comes from the literature on ethics in clinical (medical) prac- tice rather than from research ethics, but we suggest that it has obvious appli- cation to research ethics. The term microethics was originally proposed by Komesaroff (1995) to address what he saw as the disjuncture between “big issue” bioethics and everyday ethics in clinical practice. Bioethics has tradi- tionally focused on what are now the classic “neon light” debates centered on such issues as euthanasia, cloning, reproductive technologies, and so on. Bio- ethics, especially in its mainstream form, where it arises out of analytic moral philosophy, provides philosophical tools of analysis to understand the funda- mental ethical issues at stake in these debates and goes on to logically argue

Guillemin, Gillam / “ETHICALLY IMPORTANT MOMENTS” 265

toward conclusive statements about what should be done in dealing with these issues. The debates are usually conducted in general rather than partic- ular terms and may readily be interpreted as recommendations for public policy rather than as a guide to individual ethical reflection or action.

Although it has been a useful contributor at this general level, bioethics has been critiqued, from both within and outside its philosophical base, for either ignoring or being unable to deal with the ethical questions at the local or individual level. The increased focus on what is now often called “clinical ethics” or “medical ethics” has gone only part way toward dealing with this criticism. Although clinical ethics addresses itself to individual doctors (and to a lesser extent, other health professionals) on matters that arise in the clini- cal setting, the focus has still tended to be on “dilemma-type” issues, such as refusal of treatment. Clinical ethics has still been couched in general and ab- stract terms, as witnessed by the continuing tendency to cast all communica- tion between doctors and patients in terms of informed consent. Komesaroff (1995) addressed all these sorts of concerns when he claimed that

medical ethics is not just about the dramatic questions that are discussed widely in the popular media or in the philosophical texts. Ethics is what happens in every interaction between every doctor and every patient. (p. 68)

By using the term microethics, Komesaroff attempted to capture the everyday ethical issues that arise in clinical practice—the establishment of trust be- tween doctor and patient, the taking of a sexual history, the dealing with past fears, the probing about the patient’s illness experience. None of these pre- sents a “dilemma” in the classic sense that we described earlier, but Komesaroff wanted to both validate them as important ethical matters worthy of the clini- cian’s attention and also provide a language for reflecting on them.

For Komesaroff, microethics is about the complex dynamics between doc- tor and patient in clinical practice. However, it is also potentially applicable to the complex dynamics between researcher and participant in research.2 To talk about microethics in research practice is to give credence to the “ethically important moments” we articulated earlier: what to say in response to Sonia’s disclosure of sexual abuse, deciding how much to probe a participant about a difficult experience, and so on. Although the term microethics has been critiqued for setting up a false dichotomy between the abstractness of bio- ethics and the specificity of ethics in clinical practice, it nonetheless provides a language to describe and understand the “ethics in practice” in research.

The Relationship Between Procedural Ethics and Microethics

Earlier, we suggested that for many qualitative researchers, the research ethics committee application is primarily a hurdle to surmount. Being made

266 QUALITATIVE INQUIRY / April 2004

to jump this hurdle can be a frustrating experience for qualitative social re- searchers, as a recent discussion paper produced by the American Associa- tion of University Professors (2000) attests. The discussion paper documents concerns expressed by a wide range of social researchers that institutional review boards in the United States are inappropriately applying models from the biomedical sciences to ethical review of social sciences research. The dis- cussion paper reports that in the view of these social researchers, the ap- proach taken by the institutional review boards “was established and has evolved within a clinical and biomedical framework that does not fit their research” (American Association of University Professors, 2000, section I, p. 3). In this context, it is worthwhile to reflect further on the role of research ethics committee approval. Is it really the case that research ethics at the pro- cedural level does not fit the practice of qualitative research? Does the process of applying for approval in any way contribute to ethical research practice?

It is indeed true that the principles and processes used to conduct the ethi- cal review of qualitative social research were developed initially in the con- text of biomedical research. The beginnings of procedural ethics are usually traced to the Nuremberg Trials that occurred after World War II. Among those tried at Nuremberg were Nazi doctors who had committed terrible abuses on concentration camp inmates in the name of medical research. One of the out- comes of the trials was the so-called Nuremberg Code that expressly stated the obligation of medical researchers to gain the consent of those on whom they conducted research and not to harm them. The Nuremberg Code was soon followed by the World Medical Association’s Declaration of Helsinki that restated similar principles. This remains an extremely important interna- tional reference point for the regulation of medical research, as witnessed by the recent heated international debate over proposed changes to the Declara- tion that would have reduced standards of care required for clinical trials in developing countries (for a discussion of this, see Zion, Gillam, & Loff, 2000). After these international documents were drafted, individual countries be- gan to produce and revise their own codes and guidelines, which were gener- ally modeled on the Declaration of Helsinki.3

Initially, these codes and guidelines were aimed only at biomedical re- search, but in many countries they came gradually to be extended to all research involving humans, including social and qualitative research of all kinds. In the United States in the 1960s, guidelines issued by the Department of Health and Human Services through the Office of Protection from Research Risks were initially cast in broad terms to apply in principle to “all research involving human subjects” (Federal Policy for the Protection of Human Sub- jects [hereafter, the U.S. Common Rule], 2001, 46.101); however, this applied only to research that was funded or conducted by 17 federal government departments. There was also a list of exclusions regarding the types of research that did not require ethical approval. Many types of social research appeared on this list of exclusions, including surveys, interviews, and obser-

Guillemin, Gillam / “ETHICALLY IMPORTANT MOMENTS” 267

vation of public behavior (U.S. Common Rule, 2001, 46.101[b]). However, most individual institutions voluntarily extended ethical review in various ways, such as requiring review of nonfunded research as well as funded re- search and requiring researchers who believe their work is exempt to apply for exemption, a process that actually requires as much documentation and review as applying for approval (American Association of University Profes- sors, 2000, section I, p. 3). Hence, in many universities in the United States, all qualitative research is subject to review at the level of procedural ethics.

A very similar process occurred in Australia at the instigation of the National Health and Medical Research Council—the key national research funding body and organization that issues research ethics guidelines—rather than at the initiative of the universities. Although the National Health and Medical Research Council guidelines were initially directed only at biomedi- cal research, the Council issued a directive stating that it would fund research only at institutions where all research on humans was conditional on ethical review. Australian universities therefore complied, and social research of all kinds became subject to procedural ethics. In 1999, the substantially revised version of the guidelines, the National Statement on Ethical Conduct of Research Involving Humans (NS) (Commonwealth of Australia, 1999), formalized this by making the extension to all types of research explicit (McNeill, 1993, pp. 74- 75). Likewise in Canada, The Tri-Council Policy Statement (TCPS) (Medical Research Council of Canada, Natural Sciences and Engineering Research Council of Canada, & Social Sciences and Humanities Research Council of Canada, 1998) covers all research, including qualitative research. In short, in the United States, Canada, and Australia,4 a single ethics committee, often based at a university, reviews all types of research involving humans (unless subcommittees are set up for institutional convenience, but the same guide- lines still apply). All social research, whether or not it is health related, whether or not it is conducted at a health care facility, is conditional on this process.

It appears that ethics at the procedural level has been imposed on qualita- tive research from outside. Does this foreignness mean that it is irrelevant to qualitative research or worse, destructive to it? We argue that it does not. First, research ethics committees satisfy an obvious need to protect the basic rights and safety of research participants from obvious forms of abuse. We do not dispute this and would strongly support the role and importance of research ethics committees in the research process in this regard. Second, it can at least be said that procedural ethics offers researchers an ethics “checklist” by reminding the researcher to consider such issues as the potential risks to par- ticipants, the balancing of the benefits of the research against those risks, the steps needed to ensure confidentiality of data, and the inclusion of consent forms and plain language statements in the material provided to participants. This is a helpful aid in designing a research project that will be ethically acceptable in its broad methodology. Further, in fulfilling the procedural obli-

268 QUALITATIVE INQUIRY / April 2004

gations of this ethics checklist, the researcher is also granted institutional credibility to carry out the research.

However, this still leaves quite a gulf between procedural ethics and “eth- ics in practice.” The checklist is not much help once the researcher is out in the field and dealing with the realities of research practice, such as Sonia’s disclo- sure about the abuse of her daughter. So what is the ultimate significance of research ethics committee approval? By gaining approval of a committee, do researchers conduct their research in a more ethical way than they would have if they had not gained approval? There is no direct or necessary relation- ship between ethics committee approval of a research project and what actu- ally happens when the research is undertaken. The committee does not have direct control over what the researcher actually does. Ultimately, responsibil- ity falls back to the researchers themselves—they are the ones on whom the conduct of ethical research depends. Arguably, procedural ethics has little or no impact on the actual ethical conduct of research.

This view of the research ethics committee process as quite divorced from the ethics of research practice is common among qualitative researchers for whom, as we noted above, the ethics committee process comes out of a for- eign paradigm (this is discussed in Daly, 1996, p. xvii). We have two responses to it. First, we suggest that the gulf between procedural ethics and “ethics in practice” (microethics) is not as great as it may first appear. On the contrary, we argue that there is considerable continuity between ethical concerns at the procedural and practical levels and moreover, there would be something drastically wrong if this were not the case. Second, we suggest that at the level of “ethics in practice,” the “homegrown” notion of reflexivity actually encap- sulates and extends the concerns of procedural ethics. When the responsibil- ity for ethical conduct falls, as it must, on the researcher and not on the re- search ethics committee, there is still a framework for thinking about ethical conduct that is not utterly divorced from procedural ethics and yet is already integral to the good practice of qualitative research. We now argue for each of these claims in turn.

Continuity Between Procedural Ethics and “Ethics in Practice”

It would be naïve and mistaken for ethicists to suggest that procedural eth- ics is the whole of ethics—that ethical issues in the practice of research can be entirely covered by the ethics committee process. It is within the dimension of “ethics in practice” that the researcher’s ethical competence comes to the fore. By this we mean the researcher’s willingness to acknowledge the ethical dimension of research practice, his or her ability to actually recognize this eth- ical dimension when it comes into play, and his or her ability to think through ethical issues and respond appropriately. However, we argue that all of this is

Guillemin, Gillam / “ETHICALLY IMPORTANT MOMENTS” 269

directly related to, and enhanced by, what happens in the ethics committee process, even though it cannot be ensured or enforced by it. To make our case, we will identify the fundamental ethical values on which procedural ethics is based and attempt to show that these values are both (a) relevant to quali- tative social research and (b) not radically different from the issues and concerns that have already been identified as arising in the microethical dimension.

Although in general, ethical codes and guidelines are often not a great source of ethical enlightenment and typically contain vague motherhood statements, guidelines for research ethics tend to be much more precise. The U.S. Common Rule (2001), the Australian NS (Commonwealth of Australia, 1999), and the Canadian TCPS (Medical Research Council of Canada et al., 1998) all provide reasonable summary statements of the ethical principles that are central in the standard approaches to research ethics in the discipline of bioethics. In the U.S. Common Rule (46.111[a]), these are in the form of a list of criteria that must be met for ethics committee approval, which includes minimizing harm, informed consent, and protection of privacy. In the Austra- lian NS, a general principle is enunciated. The NS says that its “guiding value” is integrity, which it defines as a commitment to the search for knowl- edge and to honest and ethical conduct (Commonwealth of Australia, 1999, 1.1, p. 11). The NS’s “guiding ethical principle” is respect for persons (Com- monwealth of Australia, 1999, 1.2, p. 11). This includes respect for the auton- omy of individuals, achieved mainly by the mechanism of informed consent (Commonwealth of Australia, 1999, 1.7, p. 12); respect for privacy, achieved at least partly by rules relating to confidentiality and secure storage of data (Commonwealth of Australia, 1999, 1.19, p. 13); and respect for the dignity of persons (Commonwealth of Australia, 1999, 1.4, p. 11). The TCPS includes a very similar set of guiding ethical principles: respect for persons, respect for free and informed consent, and respect for privacy and confidentiality (Medical Research Council of Canada et al., 1998, i.5-6).

The principle of beneficence, which refers to the obligation to act in ways that benefit other people, or at least in ways that do not harm them, is high- lighted in all three documents (U.S. Common Rule, 2001, 46.111[a][1]-[2]; Commonwealth of Australia, 1999, 1.3, p. 11; Medical Research Council of Canada et al., 1998, i.6). Likewise, justice is identified in all of the documents as an important principle (U.S. Common Rule, 2001, 46.111[a][3]; Common- wealth of Australia, 1999, 1.5, p. 11; Medical Research Council of Canada et al., 1998, i.6). Justice here refers specifically to distributive justice, which in the context of research on humans mainly points to the importance of a fair shar- ing of the burdens and benefits of research across the whole community.

Among all these principles, particular precedence is given to the principle of respect for persons: the NS states that “respect for the dignity and well- being of persons takes precedence over expected benefits to knowledge” (Commonwealth of Australia, 1999, 1.4, p. 11), whereas the TCPS claims that

270 QUALITATIVE INQUIRY / April 2004

“the welfare and integrity of the individual remain paramount in human research” (Medical Research Council of Canada et al., 1998, i.5). This is quite standard in guidelines for research ethics and can also be seen in the Declara- tion of Helsinki. This is of prime importance in understanding what might be called the “ethical environment” of research and needs further exploration.

Research involving human participants starts from a position of ethical tension. In the great majority of cases, research involving humans is a process of asking people to take part in, or undergo, procedures that they have not actively sought out or requested, and that are not intended solely or even pri- marily for their direct benefit, although in some cases participants may indi- rectly benefit from the process. From the perspective of bioethics and moral philosophy, this is ethically problematic, because it fails to accord to individu- als the respect that they are owed. One way of explaining the problem is that it involves a violation of the Kantian maxim that people should never be used merely as a means to someone else’s end (that is, as a tool to achieve someone else’s aim). Another way of explaining this, derived from Kant’s emphasis on the importance of the individual as the decision maker for his or her own life, is to say that it does not adequately respect the autonomy of the individual— that it is an individual’s right to act according to his or her own purposes rather than the purposes of others.5 This tension can be resolved, however, if the subjects of the research take up the goals of the research as their own; they are then not being used as mere means or tools by the researchers. In other words, in making the research their own project jointly with the researchers, they become participants in the research rather than subjects. In practice, the standard way in which this is seen to be achieved is to obtain free and in- formed consent from participants rather than simply conscripting them. It is because of this fundamental ethical principle that informed consent is so prominent in procedural ethics and why research ethics committees spend so much time being concerned about plain language statements and consent forms. Given that these are the fundamental ethics principles in procedural research ethics, let us now consider what application or relevance they have to qualitative social research and in particular, whether they relate to the pro- cedural dimension only or also to the practical.

The inherent ethical tension identified above exists in qualitative social re- search just as it does in biomedical research. Kellehear (1996) made this point quite sharply when talking about “delicate situations” in qualitative research:

The interview is the creation of an unnatural social situation, introduced by a researcher, for the purpose of polite interrogation. It is this situation, delicate by definition, which is ethically questionable. (p. 98)

With the exception of some action research and community-based research, qualitative research is not initiated or actively sought out by its participants, nor is it aimed solely at benefiting them, although as we have suggested, there may be benefits, both direct and indirect.

Guillemin, Gillam / “ETHICALLY IMPORTANT MOMENTS” 271

Qualitative social researchers are on the whole very sensitive to this ethical tension. This recognition means that respect for autonomy and the process of informed consent is just as relevant for qualitative research as it is for biomed- ical research. Satisfying an ethics committee’s requirement for informed con- sent is not merely jumping through a bureaucratic hoop. Of course, there may be significant differences in the way that informed consent is understood as a formal process. Procedural ethics in the biomedical model work on the basis of formal written consent, complete with quasi-legal documents and signa- tures of witnesses. This is particularly emphasized in the U. S. Common Rule (2001, 46.117). However, there is nothing in the concept of informed consent that requires any of this. Signed consent forms do not constitute informed consent, they merely provide evidence (perhaps of questionable value) that consent has been given, and insofar as procedural ethics requires that consent not only be given but also documented, this documenting can occur in a num- ber of ways, such as taping a verbal consent (which is explicitly allowed for by the Australian NS; Commonwealth of Australia, 1999, 1.7). Informed consent is at heart an interpersonal process between researcher and participant, where the prospective participant comes to an understanding of what the research project is about and what participation would involve and makes his or her own free decision about whether, and on what terms, to participate.

Together with respect for autonomy, the other ethical principles enunci- ated in these documents are also highly relevant to qualitative social research. Avoiding causing harm to participants, for example, is surely an absolutely basic consideration. Harm, as is widely recognized, is not merely a matter of physical harm but has many other aspects, including emotional and social harm.6 Hence, harm can potentially be caused not only by drugs and invasive procedures in medical research but also by the research methods, such as par- ticipant observation or in-depth interviewing, typically used in qualitative research. The possibility of harm to participants is the reason that research ethics committees are concerned about the methodology of research projects and the skills of researchers. The committee needs to satisfy itself that partici- pants are not being exposed to excessive or unnecessary risks, which may well be the case if the methodology of a project is not sound or a researcher lacks essential skills.

Although most qualitative researchers are well aware of issues of harm, the language in which this idea is often couched in ethics application forms may appear at best ambiguous or at worst, completely irrelevant to qualita- tive research. The potential harms to participants in qualitative social research are often quite subtle and stem from the nature of the interaction between researcher and participant. As such, they are hard to specify, predict, and describe in ways that ethics application forms ask for and likewise, strate- gies for minimizing risk are hard to spell out. In Sonia’s case, for example, the researcher asking more questions than Sonia feels comfortable answering may cause harm; or harm may be caused by not showing enough interest so

272 QUALITATIVE INQUIRY / April 2004

that Sonia feels she has been ignored and disregarded. If the researcher encourages her to seek outside help or report the situation to the police, Sonia may feel guilty for not having done this before or feel that the researcher is blaming her for being too passive, or she may be frightened by the prospect of outside intervention. Procedural ethics is obviously not the forum in which issues of potential harm and other such “ethically important moments” can be fully dealt with—they are too specific and nuanced. Again, it is at the level of “ethics in practice” that researchers must do the real ethical work in this regard. However, research ethics committees perform an important function at the procedural level by forcing qualitative researchers to reflect and think about the potential harms of their research, even though the questions may not be answerable at this level. Hence, we concur with the claim in the TCPS that these principles are “ethical norms that transcend discipline bound- aries. . . . Fundamental ethical issues and principles in research in human sub- jects are common across the social sciences and humanities . . . and health sci- ences” (Medical Research Council of Canada et al., 1998, i.2). However, we emphasize that these ethical principles alone cannot be expected to address the everyday ethical difficulties of research practice.

REFLEXIVITY AND “ETHICS IN PRACTICE”

As we have argued, ethical research is much more than research that has gained the approval of a research ethics committee. Although research ethics committees do play an important role in highlighting ethical principles that are relevant to, and important for, social research, their role is necessarily lim- ited. Research ethics committees cannot help when you are in the field and difficult, unexpected situations arise, when you are forced to make imme- diate decisions about ethical concerns, or when information is revealed that suggests you or your participants are at risk. We suggested earlier that micro- ethics might provide a discursive tool to articulate and to validate the kinds of ethical issues that confront researchers on a day-to-day basis. However, microethics is not helpful in addressing and dealing with these issues when they arise. We need a process and a way of thinking that will actually lead to ethical research practice. This is where we see an important role for reflexivity.

Reflexivity and Research

Reflexivity is a concept very much at home in the world of qualitative social research. We want to argue that it is also a bridge to the procedural ethi- cal issues that can often seem out of place in the everyday practice of social research. Reflexivity, we suggest, is closely connected with the ethical prac-

Guillemin, Gillam / “ETHICALLY IMPORTANT MOMENTS” 273

tice of research and comes into play in the field, where research ethics com- mittees are not accessible.

Research is primarily an enterprise of knowledge construction. The re- searcher (and coresearchers), with his or her participants, is engaged in pro- ducing knowledge. This is an active process that requires scrutiny, reflection, and interrogation of the data, the researcher, the participants, and the context that they inhabit. We speak of research as a reflexive process, but how do we define this slippery concept? Mason (1996) stated that reflexive research

means that the researcher should constantly take stock of their actions and their role in the research process and subject these to the same critical scrutiny as the rest of their “data.” (p. 6)

Hertz (1997, p. viii) noted that the reflexive researcher does not merely report the “facts” of the research but also actively constructs interpretations (“What do I know?”), while at the same time questioning how those interpretations came about (“How do I know what I know?”).

Jenkins (1992) observed how Bourdieu provided another helpful way of thinking about reflexivity in research. Bourdieu suggested the reflexive pro- cess comprises taking two steps back from the subject of the research. The first step back is the objective observation of the research subject; the next step back is the reflection of the observation itself. This is akin to the first step pos- ing the “What do I know?” question and the second step asking the “How do I know?” question.

Reflexivity in research is not a single or universal entity but a process—an active, ongoing process that saturates every stage of the research. Harding (1986, 1987, 1991) reminded us that as researchers, our social and political locations affect our research. Our research interests and the research ques- tions we pose, as well as the questions we discard, reveal something about who we are. Our choice of research design, the research methodology, and the theoretical framework that informs our research are governed by our values and reciprocally, help to shape these values. Who we include and who we ex- clude as participants in our research are revealing. Moreover, our interpreta- tions and analyses, and how we choose to present our findings, together with whom we make our findings available to, are all constitutive of reflexive research. Reflexivity in research is thus a process of critical reflection both on the kind of knowledge produced from research and how that knowledge is generated.

Using Reflexivity in Research

Having acknowledged the construction of knowledge in research as a reflexive process, how does this assist us in dealing with the microethics of research practice? Reflexivity is not usually seen as connected with ethics at

274 QUALITATIVE INQUIRY / April 2004

all. Rather, reflexivity in qualitative research is usually perceived as a way of ensuring rigor (Finlay, 1998; Koch & Harrington, 1998; Rice & Ezzy, 1999). Reflexivity involves critical reflection of how the researcher constructs knowledge from the research process—what sorts of factors influence the re- searcher’s construction of knowledge and how these influences are revealed in the planning, conduct, and writing up of the research. A reflexive re- searcher is one who is aware of all these potential influences and is able to step back and take a critical look at his or her own role in the research process. The goal of being reflexive in this sense has to do with improving the quality and validity of the research and recognizing the limitations of the knowledge that is produced, thus leading to more rigorous research. It does not have an overtly ethical purpose or underpinning.

To see how a useful connection can be made between reflexivity and eth- ics, consider again the accounts of reflexivity that we offered. These accounts suggest that reflexivity is not necessarily focused only on the production of knowledge in research (what might be called the epistemological aspect of research practice) but also on the research process as a whole. Adopting a reflexive research process means a continuous process of critical scrutiny and interpretation, not just in relation to the research methods and the data but also to the researcher, participants, and the research context.

This would include issues about the ultimate purpose of the research. Is the aim to construct knowledge, advance the researcher’s career, further the specific goals of the research participants, or some more theoretically based purpose, such as “enabling community life to prosper” (Christians, 2001, p. 145)—and is this an ethically appropriate purpose? It would also include the interpersonal aspects of research, the interactions between researcher and participant that we suggest, are the substrate of the ethical dimensions of research practice. In these interactions lie the possibilities of respecting the autonomy, dignity, and privacy of research participants and also the risks of failing to do so, thus perhaps causing harm to the participants in various ways. It is in these interactions that the process of informed consent really occurs—not on the pieces of paper that an ethics committee peruses. It is therefore in these interactions that the integrity of the researcher is really on the line.

In being reflexive in this sense, a researcher would be alert not only to issues related to knowledge creation but also ethical issues in research. This alertness might include conscious consideration of a range of formal ethical positions and adoption of a particular ethical stance. For example, a feminist communitarian position implies that research be undertaken in a particular way. In this model, as Christians (2000, p. 145) described, research partici- pants have a say in how the research is framed, what questions are asked, what methods are used to answer them, and how the findings are regarded and used. This is a particularly strong and proactive way of respecting auton- omy, going far beyond the rather minimalist notion of informed consent.

Guillemin, Gillam / “ETHICALLY IMPORTANT MOMENTS” 275

However, a simple decision to adopt such a position would not necessarily lead to ethical research practice; ongoing reflexivity on the part of the re- searcher is still required. This is partly to check that the researcher’s practice is actually embodying his or her principles; in addition, this allows the re- searcher to become aware of situations where following the theoretical posi- tion may not be the best course and may not best uphold the interests of his or her participants. For instance, some research participants may not want to take the very active role envisaged for them in the feminist communitarian model, and reflexivity is needed to notice this and respond in an ethical way. Out of this sort of reflexivity, it is possible that new or modified theoretical positions may grow, but the key point is that the ethical practice of research requires more than the in-principle adoption of an enlightened ethical theory.

McGraw, Zvonkovic, and Walker (2000) have pointed to the connection between reflexivity and ethics. In an article that focuses on a practical ethical problem that they encountered in a research project, McGraw et al. discussed ethics as one aspect of reflexivity:

[Reflexivity is] a process whereby researchers place themselves and their prac- tices under scrutiny, acknowledging the ethical dilemmas that permeate the research process and impinge on the creation of knowledge. (p. 68)

McGraw et al. did not develop this idea, but the comment suggests that our approach is not stretching the concept of reflexivity beyond its bounds.

On this basis, we can say a little more about how reflexivity could contrib- ute to ethical research practice. Being reflexive about research practice means a number of things: first, an acknowledgment of microethics, that is, of the ethical dimensions of ordinary, everyday research practice; second, sensitiv- ity to what we call the “ethically important moments” in research practice, in all their particularities; and third, having or being able to develop a means of addressing and responding to ethical concerns if and when they arise in the research (which might well include a way of preempting potential ethical problems before they take hold).

For example, with the opening scenario of Sonia, a reflexive researcher would first understand and acknowledge that ethics in research has different dimensions and that research practice is infused with microethics at all levels. Second, this understanding would lead to an awareness of the potential im- pact of the researcher and the research process on the lives of Sonia and the other participants before the research even commences. This reflexive ques- tioning would consider the impact of the researcher posing the questions to Sonia, the questions and wording used by the researcher, the way the ques- tions are framed, as well as the timing of the questions. This reflexive process cannot predict all likely ethical problems that may arise in the doing of research. However, what it does offer is a process that could have foreseen at least the general sort of predicament that arose in relation to Sonia, or other potential sorts of consequences that might arise, and offered the researcher

276 QUALITATIVE INQUIRY / April 2004

ways of addressing them. This may suggest ways of responding to Sonia, including whether to encourage her to talk about the suspected abuse, whether to turn off the interview tape, whether to offer Sonia assistance and the type of assistance offered, and whether to continue with the interview after her disclosure. These issues require decisions that are difficult to make without forethought and prior consideration. They are also difficult to make well without an ongoing awareness of the impact that a researcher’s re- sponses can have on a research participant.

Reflexivity is not prescriptive in the sense that it specifies in the abstract precisely what a researcher should do in response to Sonia’s situation. How- ever, it does have a number of ethically important functions. In being reflex- ive, researchers both reflect about how their research intervention might affect the research participants before any actual research is conducted and consider how they would respond as a researcher in the sorts of situations that they can at this stage only envisage. Leading on from this, reflexivity encourages researchers to develop the skills to respond appropriately. In the actual conduct of research, the reflexive researcher will be better placed to be aware of ethically important moments as they arise and will have a basis for responding in a way that is likely to be ethically appropriate, even with unforeseen situations. Our notion of reflexivity urges researchers to be reflex- ive in relation to interpersonal and ethical aspects of research practice, not just the epistemological aspects of rigorous research.

CONCLUSION

Our focus in this article is an examination of what constitutes ethics in qualitative research. Starting off from a position of seeing procedural ethics and “ethics in practice” as divorced from one another, we suggest that there is some common useful ground to consider. Although procedural ethics is unable to inform and guide all aspects of research practice, it does serve a valuable function in forcing us to consider and reflect on the fundamental guiding principles that govern research integrity. Furthermore, it acts as a practical reminder that we need to be both mindful and active in protecting our research participants (and ourselves) from harm and undue risks, as well as affording respect for autonomy.

The concept of microethics is a valuable discursive tool to allow us to talk about, validate, and better understand the ethically important moments in research practice. It is limited, however, when it comes to offering guidance in how to deal with these ethical events. We have argued that this is where the notion of reflexivity is most useful. Reflexivity is something that most quali- tative researchers are aware of and incorporate into their research practice; it is not a new concept. However, what we suggest is that its meaning be ex- panded so that reflexivity be considered and enacted as a way of ensuring not

Guillemin, Gillam / “ETHICALLY IMPORTANT MOMENTS” 277

just rigorous research practice but also ethical research practice. Being reflex- ive in an ethical sense means acknowledging and being sensitized to the microethical dimensions of research practice and in doing so, being alert to and prepared for ways of dealing with the ethical tensions that arise. As we have stated, reflexivity does not prescribe specific types of responses to research situations; rather, it is a sensitizing notion that can enable ethical practice to occur in the complexity and richness of social research.

NOTES

1. Research ethics committees or boards are known by different names in different countries; for example, in the United States these committees are known as institutional review boards, in Australia they are human research ethics committees, in the United Kingdom they are local research ethics committees, and so on. For simplicity, we will use the generic term research ethics committees to refer to all bodies that review applica- tions for ethical approval of research involving humans.

2. Komesaroff (1996, p. 46) noted this, but does not expand on it in this text. 3. McNeill (1993) provided a comprehensive account of the development of the

ethics committee process in a number of different countries. 4. The United Kingdom, however, is somewhat different in that local research ethics

committees are still largely the province of the health care system and are attached to hospitals or local health authorities and hence, deal primarily with medical research. Ramcharan and Cutcliffe (2001) discussed the current problems in the United Kingdom of assessing social research within the largely medical research ethics committee framework.

5. This Kantian perspective is especially evident in the Tri-Council Policy Statement (Medical Research Council of Canada, Natural Sciences and Engineering Research Council of Canada, & Social Sciences and Humanities Research Council of Canada, 1998) that refers frequently to the notion of using a person as a means.

6. A standard definition of harm in bioethics is that it is “a setback to a person’s inter- ests.” What counts as interest is controversial, but it clearly extends beyond an interest in physical well-being. This controversy regards whether interests is a subjective or an objective notion, but this means that interests can conceivably extend to include any aspect of a person’s life that that person regards as important. Hence, there are many ways in which a person can be harmed.

REFERENCES

American Association of University Professors. (2000). Protecting human beings: Institu- tional review boards and social science research. Retrieved December 13, 2002, from http://www.aaup.org/statements/Redbook/repirb.htm

Bouma, G., & Diemer, K. (1996). Human ethics review and social sciences: Several emerging issues. Monash Bioethics Review, 15(1), 10-11.

Bulmer, M. (Ed.). (1982). Social research ethics. London: Macmillan.

278 QUALITATIVE INQUIRY / April 2004

Christians, C. (2000). Ethics and politics in qualitative research. In N. Denzin & Y. Lincoln (Eds.), Handbook of qualitative research (2nd ed., pp. 133-155). Thousand Oaks, CA: Sage.

Coady, M., & Bloch, S. (Eds.). (1996). Codes of ethics and the professions. Melbourne, Aus- tralia: Melbourne University Press.

Commonwealth of Australia. (1999). National statement on ethical conduct in research involving humans (Cat. No. 9818566). Canberra: Australian Government Publishing Service.

Daly, J. (Ed.). (1996). Ethical intersections: Health research, methods and researcher responsi- bility. Sydney, Australia: Allen & Unwin.

Federal Policy for the Protection of Human Subjects, 45 C.F.R. § 46 (2001). Retrieved January 3, 2003, from http://ohrp.osophs.dhhs.gov/humansubjects/guidance/ 45cfr46.htm

Finlay, L. (1998). Reflexivity: An essential component for all research? British Journal of Occupational Therapy, 61(10), 453-456.

Harding, S. (1986). The science question in feminism. Ithaca, NY: Cornell University Press. Harding, S. (Ed.). (1987). Feminism and methodology. Bloomington: Indiana University

Press. Harding, S. (1991). Whose science? Whose knowledge? Thinking from women’s lives. Milton

Keynes, UK: Open University Press. Hertz, R. (1997). Introduction: Reflexivity and voice. In R. Hertz (Ed.), Reflexivity and

voice (pp. vi-xviii). Thousand Oaks, CA: Sage. Homan, R. (1991). The ethics of social research. London: Longman. Hornsby-Smith, M. (1993). Gaining access. In N. Gilbert (Ed.), Researching social life

(pp. 52-67). London: Sage. Jenkins, R. (1992). Pierre Bourdieu. London: Routledge Kegan Paul. Kellehear, A. (1996). Unobtrusive methods in delicate situations. In J. Daly (Ed.), Ethical

intersections: Health research, methods and researcher responsibility (pp. 97-105). Sydney, Australia: Allen & Unwin.

Kimmel, A. (1988). Ethics and values in applied social research. Newbury Park, CA: Sage. Koch, T., & Harrington, A. (1998). Reconceptualizing rigour: The case for reflexivity.

Journal of Advanced Nursing, 28(4), 882-890. Komesaroff, P. (1995). From bioethics to microethics: Ethical debate and clinical medi-

cine. In P. Komesaroff (Ed.), Troubled bodies: Critical perspectives on postmodernism, medical ethics and the body (pp. 62-86). Melbourne, Australia: Melbourne University Press.

Komesaroff, P. (1996). Medicine and ethical conditions of modernity. In J. Daly (Ed.), Ethical intersections: Health research, methods and researcher responsibility (pp. 34-48). Sydney, Australia: Allen & Unwin.

Mason, J. (1996). Qualitative researching. London: Sage. McGraw, L., Zvonkovic, A., & Walker, A. (2000). Studying postmodern families: A femi-

nist analysis of ethical tensions in work and family research. Journal of Marriage and the Family, 62(1), 68-77.

McNeill, P. (1993). The ethics and politics of human experimentation. Cambridge, UK: Cam- bridge University Press.

Medical Research Council of Canada, Natural Sciences and Engineering Research Council of Canada, & Social Sciences and Humanities Research Council of Canada. (1998). Tri-Council policy statement: Ethical conduct for research involving humans (Cat.

Guillemin, Gillam / “ETHICALLY IMPORTANT MOMENTS” 279

No. MR 21-18/1998E). Ottawa, Canada: Public Works and Government Services Canada.

Orb, A., Eisenhauer, L., & Wynaden, D. (2001). Ethics in qualitative research. Journal of Nursing Scholarship, 33(1), 93-96.

Ramcharan, P., & Cutcliffe, J. R. (2001). Judging the ethics of qualitative research: Con- sidering the “ethics as process” model. Health and Social Care in the Community, 9(6), 358-366.

Rice, P., & Ezzy, D. (1999). Qualitative research methods: A health focus. Melbourne, Austra- lia: Oxford University Press.

Robley, L. R. (1995). The ethics of qualitative nursing research. Journal of Professional Nursing, 11(1), 45-48.

Zion, D., Gillam, L., & Loff, B. (2000). The Declaration of Helsinki, CIOMS and ethics of research on vulnerable populations. Nature Medicine, 6(6), 615-617.

Marilys Guillemin is a senior lecturer at the Centre for the Study of Health and Society at the University of Melbourne, Australia, where she teaches health sociology and qualitative research methods to medical students and postgradu- ate social health students. She has published widely for academic, health profes- sional, and lay community audiences. Her research work broadly covers the sociology of health and technology and women’s health in particular.

Lynn Gillam is a senior lecturer in health ethics at the Centre for the Study of Health and Society and a research fellow in the Research Centre for Applied Philosophy and Public Ethics at the University of Melbourne, Australia. She has published in international journals, including Bioethics, Journal of Medical Ethics and Medicine and Philosophy, and is the author of a well- known series of introductions to topics in bioethics published in Monash Bioethics Review. She is coauthor of Facts and Values: An Introduction to Critical Thinking for Nurses (MacLennan & Petty, 1995), which includes a large section on ethical reasoning and decision making.

280 QUALITATIVE INQUIRY / April 2004