Some practical notes on ethics applications

Alright, so the first two blog entries on ethics weren’t very fun (here and here). I’ll admit that. And it’s a danger when dealing with a topic like university human research ethics review that we may contribute to the sense students (and others) have that it’s a dry or dreadful subject. I worry that we tell our horror stories to our students and prepare them for the worst from the ethics process, forgetting that this can set up self-fulfilling expectations. I didn’t help that with the last couple of posts.

So, in the interest of punching up the entertainment value (after all, I have to compete with some brilliant posts on the new Creationist museum in Missouri, my home state, by Lisa Wynn and ongoing cultural observations from Nursel and Jovan), I’m going to take a little different strategy and write in a more conversational tone.

Although the institutional dynamics of something like the Ethics Review Committee (Human Research) at Macquarie might be fascinating to a few (mostly to me), the majority of our readers at Culture Matters are likely to be more interested in practical questions. So I’ll try to highlight the most important, recurring issues for ethnographic projects from my perspective as researcher, ethics advisor, and application reviewer on the committee. Although all of the examples I discuss below are fictional, some resemblance to individuals living or deceased is inevitable. But please know, if it sounds like I’m talking about you, and you’re a Macquarie student or faculty member, you’re probably part of fictional synthesis because none of these issues is rare or unusual.

The over-arching issue for anthropologists proposing ethnographic fieldwork (IMHO) is probably that we don’t spend enough time nutting out the pragmatics of fieldwork before we go. We often just assume that students will figure it out on their own. While this may have been an effective strategy at one point and at some places, the short time horizon for research degrees (our MAA and PhD) makes this strategy less likely to succeed. The contemporary realities of ethics review also mean that now, perhaps more than in some earlier periods, we need to think about research methods before we enter the field.

Ten biggest issues in no particular order:

#1: Applications contradict themselves internally in several places. This is not just a problem for anthropologists; for example, an applicant says she’s going to get permission to record or photograph her subjects, but fails to include any mention of this in the information and consent procedure. Or a student says he’s working in one place, but has letters of support from another. (I know, you were saying to yourself, ‘maybe they won’t notice…’, but you forget, committee members get really good at reading these forms…)

Admittedly, some committee members, even at Macquarie, can get overly finicky about these sorts of inconsistencies, but one of the concerns is that such inconsistencies are a sign that the researcher hasn’t really thought about things. Best advice: read quickly through the form before setting off to work and check yourself as you go to make sure that you’re being consistent. One of the places on the Macquarie form that frequently causes trouble is item 6.6 on giving subjects feedback. Frequently, the answer here has nothing to do with the actual information given to participants; in Macquarie’s Department of Anthropology, we’re setting up on on-line clearing house where students can circulate for their subjects ‘research reports,’ short, non-jargon-filled accounts of their findings. It’s a great way to publicize the kinds of research done in the department (without foreclosing other forms of publication) and, at the same time, create an easy channel to give something back to the subjects generous enough to work with us.

#2: No interview questions. I say this baldly even though I know that interview questions are a controversial demand. My approach is that, in most projects, some sort of non-binding, sample interview questions can and should be provided. Some projects — like life histories — make these sorts of provisions less appropriate, and I’ll typically argue against such demands in committee meetings if other members bring them up.

HOWEVER, many ethnographers say that they are going to do ‘in depth’ interviews on very specific topics, as much of contemporary anthropological inquiry is topic-based rather than holistic research in small-scale communities. In this case, some sense of the kinds of questions you might ask helps the committee to understand the invasiveness, sensitivity, or extensiveness of the interview process that the applicant proposes. This is especially important if your research topic is controversial or sensitive, less critical if it’s innocuous (not an insult; my research probably falls under the latter category). But I want to emphasize this: the questions are not binding IF the applicant makes it clear that the interview format is open, non-invasive, and not overly personal. At least at Macquarie, the ethics reviewers know what this means.

What seems to me to be the most common situation is that ethnographers haven’t really thought that much about the pragmatics of interviewing or they’ve become so convinced that they are self-reflexive, auto-deconstructing, and co-creational in their research that they are ‘politically opposed’ to asking questions. Then how the hell are you going to conduct interviews? If you’re really not doing interviews — that is, you’re just going to do participant observation — then don’t say you’re doing interviews on the application. Interviews imply questions, at least some sense of what you will ask about.

What I tend to ask students is, if you get off the plane, catch a bus into town or to your site, introduce yourself to an important member of the community, explain your project, and the person says, unexpectedly, ‘Great, I want to help. Turn on your recorder. What do you want to know?’ what are you going to ask? Are you really going to have no questions? According to one of my former professors at the University of Chicago, the late David Schneider used to ask students a similar question in the proposal defenses for doctoral research: ‘What are you going to do when you get off the plane (or boat or bus or train…)?’ The question would frequently stop pre-field students in their tracks. I’ve had similar halting conversations with pre-field students. So one thing that I suspect is that the ethics application may be the first time that novice researchers have ever even thought this practically and concretely about field methods in ethnography (note: I know other fields, probably even other anthropology departments, prepare their students differently. This applies to a limited number of places and types of projects. We’re trying to prepare students here better for things like interviewing).

#3: Changes in research: One of the issues raised by Katz’s article on ethics review is the possibility — nay, inevitability — that research agendas change over the course of an ethnographic project. How can you ask for prior approval if the research agenda and methods are themselves emergent? The easy answer to that (and I’m not sure that the whole committee here at Macquarie would agree with me) is that you do the best you can and then remember that ethics review has a dialogic and ongoing quality in ethnography, especially long-term fieldwork. Some changes seem to warrant serious consultation: for example, if a hot-button of risk, working with children, learning of illegal activity, etc. (see below) comes up in research where it was not anticipated.

If the subject matter changes and crosses into areas where significant NEW ethical issues are raised, then I recommend contacting the ethics administrators. We have excellent ones at Macquarie: Kokila de Silva (a former student in our department), Fran Thorpe, and Nicola Meyton among them (research office staff here but you can also use their general email, ethics.secretariat@vc.mq.edu.au. Other advisors are available in every division of the university (the list is here). Whoever you contact is most likely to run any changes by the chair of the committee and some members that are particularly well versed in the area you’re working on. You’re likely to be given some advice and feedback but virtually never told ‘you can’t follow this research.’ If the change is very significant, we’ll probably just ask you to submit an ‘amendment’ to the ethics form (that is, a new version of whichever items of the application have changed). For example, if your research starts to uncover a lot of illegal activity, and you’ve decided to directly investigate it, we’d ask you to submit a new version of item 3.8.

You won’t be ‘in trouble,’ but the Ethics Review Committee will want to make sure that you’ve thought about how to handle it and put some safeguards in place; we might be able to offer some advice (you’d be surprised — we read some really clever ways of dealing with issues like this), and we’ll be able to back you up later if there are questions about this. For example, in Australia, you are legally bound to report to police certain sorts of crimes (for example, abuse of children). Many researchers may not realize this; although you’re unlikely to get prosecuted, you may find yourself in a very uncomfortable position.

(I understand that other IRBs and ethics review committees may not handle this ongoing interaction well, and our committee does not always handle precedent-setting cases in the best possible way. But we do very much try to learn from experience and help other researchers to share the hard-won wisdom that faculty around campus are gaining. As I’ve said in an earlier blog post, we are in the business of supporting good research, not making it impossible. As Dr. Schrag and others have pointed out, some — many? — other IRBs may lose the plot on this part of the agenda.)

#4: A related recurring problem in anthropology applications is to misinterpret how to deal with the question of ‘risk.’ I take others’ comments, such as Dr. Schrag’s discussion of the ‘canard of interview trauma,’ to heart, but there is a problem with the way that we, anthropologists, tend to discuss risk in our proposals.

Some parts of our field like to play up the drama, the controversy, the conflict, the Sturm und Drang of our projects. Some anthropologists truly do amazing, harrowing research, in fieldsites and social situations where the intensity of human drama cannot be questioned. Nevertheless, an ethics application is not the place for the most hyperbolic, emotional, or explosive description of the conflict in your research; both the researcher and the committee need a pretty sober assessment of the risks involved to subjects and researcher.

If you write a research description (10.1 in the Macquarie form) that plays up the conflict in your website, the marginal legality of your subject, the possible dramatic dangers, the ‘frontline’ quality of the research, and similar traits, — or better yet, cut and paste it from another document — and then you check the box in 3.1 that says ‘no’ risks to anyone participating (this includes you), then you are asking the committee to send your application back to explain the discrepancy. Save the adrenaline-drenched prose about ethnographic research for other forums; it’s a bad idea to play this up in the ethics application.

The point is not that there are no risks in good research. Good research may include serious risks. But the way to approach those is to think about them, minimize or mitigate them, plan for foreseeable problems, or even describe why risks are worth running. For example, ethnographers are surprisingly reluctant on the Macquarie form to talk soberly about the public interest dimension of our research, the indirect effects that better knowledge might have for the good of society as a whole, on item 3.7. We have researchers at the university who go to war zones, who conduct experimental medical procedures, who ask about trauma and human rights violations and genocide and crime, who regularly work with children or the mentally disabled. Our committee does not have a problem with research that has risks; we do have a problem when the answers to questions about risk are answered in a way that suggests the researcher hasn’t thought about them.

Short suggestion: Simply checking the box to indicate that there are ‘no risks’ is not the ‘easy way’ to get through review. And the ethics application is not the place to play up the drama and danger inherent in your project; save it for your undergrad students in lecture.

(In passing, and with no expectation that this will lesson the likelihood of students doing a particular project, subjects wherein the researcher is in danger do seem to exercise a disproportionate fascination among our students. I frequently ask them if the intellectual issues that they’re interested in couldn’t be studied elsewhere, and they give me a look of disgust that would usually be reserved for someone with an indecent proposal. Alas, I’ll keep trying. But if the risks of a project are very great and the potential public good to come of it low, is it really a good project? I know, you’re getting that look even reading this…)

#5: Another way to make the ethics clearance process go off the rails is to turn the proposal in at the last minute. A really stunning number of the problem cases we confront are simply researchers who want to ram through proposals that are, at best, half cooked; our committee is actually very accommodating, and our turnaround, especially in the ‘off peak’ season, can be quite good (I think we promise 20 days, but see the committee’s website for certain). The proposals that come in at the last minute, however, invariably are the ones with missing information, no information & consent form, suggestions that an ad will be used to recruit participants but no ad provided, serious gaps in the discussion of methods, no letters of support from organizations that are said to be cooperating, etc. There’s some weird inverse law that, the less time applicants have to get approval, the more likely they are to fail to fill out the form comprehensively.

Give yourself a couple of months, especially if you have a topic that’s likely to raise any red flags (working with children, investigating illegal activity, going into conflict zones, asking very personal questions such as about medical histories, working in a place under authoritarian regime, studying human rights struggles where activists are disappearing, seeking permission to work with Aboriginal Australians…). Bottom line is that your failure to plan is not an emergency for the committee.

If you’re waiting on a research visa or a letter of support, but everything else is in order, SUBMIT immediately. The committee is likely to give you conditional approval pending the forwarding of this letter to our office; even if you have to go to the field first to get, as is often the case with ethnographic projects, we will give you approval and ask you to forward on this material once you get it. The same goes for other details, such as the specific organizations you’ll be working with, the translations of information and consent forms, and the like; if one or a couple of details is holding up the process, please submit and explain in a cover letter. You’re quite likely to get conditional approval if the rest of the application is done well. Better to submit with a missing element than to submit with no time before the start of the project.

#6. Informed consent. This is a major issue with ethnographic researchers, as postings by Rena Lederman and Alex Golub, personal discussions with other researchers have really highlighted for me, and ongoing observations as an ethics advisor make abundantly apparent. Informed consent, however, is one of the pillars of research ethics; since the Nuremberg Code in 1948, voluntary, informed consent of research subjects has been a cornerstone of human research. Although some commentators might argue that this code should apply primarily to medical research and laboratory research (such as psychology), and should not apply to non-invasive observation in a naturalistic setting, I’m not one of those.

Informed consent, however, comes in many forms; in many anthropological settings, oral consent is more appropriate, so answering ‘yes’ to item 7.1 is not compulsory (at least a quarter to a third of all projects at Macquarie check ‘no’ here, I’d estimate with no systematic review). Although the Macquarie ethics application provides a lengthy and involved template for a consent form (item 7.4. a), there’s many reasons why one might depart from this. In my experience, a substantial minority of ethnographic projects do depart from it in some way; the ones that correspond are with highly literate, Western populations who will understand consent or release forms.

My advice to students, however, is to think seriously about how you want to go about informing people and describe your ideal process to the committee. If you need to be very careful about protecting anonymity, for example, even a signed form might prove more dangerous than not getting one. If you are dealing with an illiterate or semi-literate population, or even one that would be suspicious about signing documents, then it’s probably best not to use a written consent form.

That said, read the template, think about what the committee is asking you to accomplish, and prepare a consent ‘script’ that you will use or work from. The biggest problem for many novice ethnographers is that it can be hard to communicate clearly to our collaborators what we’re trying to do in a non-technical language. Hell, I had problems communicating with my grandmother about what I was trying to do in my research!

For me, the key issues are: who are you and what are you doing? What you hope to accomplish with your research (although sometimes it’s hard to even explain a degree)? What are you asking people to do and are there any risks (if not, no need to write about it)? Do you want to record them and do you have their permission? How are you going to protect them if there’s any danger? What are they going to get from you? That, in a nutshell, is the key set of issues. If you’re dealing with a really superficial investigation or short observation or if you’re interviewing people for whom this process is going to be really alien, what I sometimes suggest is that the researcher get collaborators in the field who are better educated, who will be given more substantial information, and after the oral consent process, the researcher tell the subjects that, if they have any questions or problems, talk to the collaborator. The whole oral consent process should really be quite brief, probably less verbiage than I’m spending talking about informed consent. Too long and complicated, and it actually defeats the purpose of informing (as any lecturer knows, one way to make sure someone doesn’t understand a set of concepts is to give them more and more concepts).

For several projects, I’ve suggested a ‘tiered’ approach to information and consent; first, under no circumstances (except to avoid serious risk to the researcher temporarily) does the researcher conceal what he or she is doing. If you have to hide that your doing research, it’s a dangerous situation to be in, and I suspect one that an experienced ethnographer only should be doing. Second, if you are working with a group where its reasonable to inform people what you’re doing, try to use local channels to do so. Third, if you are actually taking person information from someone, you need to do a real oral consent process. If you’re sitting at a bus stop talking to someone about the building across the street, you don’t have to get them to sign a form. In our committee, at least, a tiered approach that differentiates public behaviour where a person has no reasonable expectation of anonymity from private behaviour or personal information through direct questioning has been upheld.

Public officials speaking in their roles and public behaviour, in which there is neither an expectation of anonymity, nor an attempt to identify subjects individually, really needs no consent, as far as I’m concerned. Although the researcher shouldn’t conceal what he or she is doing, nor try to exploit the subjects, there are many public events where informed consent is practically impossible and ethically unnecessary. In many of the capoeira events that I attended in Brazil, for example, a half-dozen people might be videotaping at any one time, and people making public speeches knew that they might be quoted by local media or anthropologists (sometimes I was not the only ethnographer present).
With long relationships, once a collaborator is clear about the researcher’s project and agenda, and has gone through the consent process, I think it’s not incumbent to keep going through the process again, but I do think that there are ways that a researcher can remind the subject of the nature of the relationship. Openly taking notes or asking if you can record something for your research, although not mandated, is one way to remind your subjects that what they say is being taken down. I find that, if I conduct myself responsibly, most collaborators have no problem with this. But I intentionally make it clear that I’m doing research, and, if they ask me to, I don’t record what I’m hearing or take notes, allowing people to talk ‘off the record.’ Of course, what they say will influence me, even if unconsciously, so it’s likely to enter the research surreptitiously; but that’s often what my subjects intended (at least in my work in Brazil).

This is probably more confusing than it needs to be, but the point is just that the ethics committee does not believe that a one-size-fits-all form will apply to all research projects. A substantial portion of the projects that we review are approved without using the ideal template, but all applicants need to think seriously about how they will fulfill the principles of seeking informed consent rather than just getting forms filled out. As always, protecting our subjects is our first priority, protecting our researchers, second, and assuring that good research with integrity is carried out is third (but still up there).

#7. Many ethnographers seem to get confused about the relationship between confidentiality and anonymity. Most ethnographic research is not strictly anonymous because the researcher typically knows the subject. However, ethnographic writing tends to protect anonymity. In a project where anonymity of subjects is a serious concern, either for safety or protection from social stigma, then the researcher should really think about how to build safeguards for confidentiality. This goes without saying.

Different countries (and probably even different states) have different laws about confidentiality for doctors, lawyers, journalists, and other professions, but none that I know of protects anthropologists and their subjects. So, if you’re taking notes on something that the government might want to know about, you won’t be able to protect those notes if they’re subpoenaed, at least not in any jurisdiction I know. The only thing worse than refusing to turn over your information in those settings, however, is to try to destroy the records, so you need to think about this ahead of time.

This said, however, I think that anthropologists’ concerns about their data falling into the wrong hands seem to me, except in certain exceptional cases, to be overblown. When I was applying for language study grants in the 1990s, some of us were worried about accepting area studies money that was clearly allocated by the government for strategic purposes; now that a lot of these funds have dried up, I’m sure I’d feel a lot less morally suspect intercepting some money from the government that might otherwise be spent on corrupt subcontracting, various pork boondoggles, or yet another tax cut. If you’re collecting this sort of information, however, it’s in your interest to learn what the rules are. Since you won’t be able to protect your notes, think of some way to make sure that they won’t be compromised (I, for example, have handwriting that I even find illegible after a very short period of time.).

But it’s also possible that, in some projects, subjects will not want to be anonymous. I think a good case can be made that public recognition is one of the things that we hope to give to our collaborators (item 3.7 in the Macquarie form, again). If you’re not sure, make it part of the consent process; a check box that allows subjects to choose. In my work on capoeira, recognition was one of the motives for people to work with me at all. Nevertheless, I still protected people if they were saying things that might reflect poorly on them or cause conflict in their home communities.

#8. Remember that your thesis IS a public document. So saying that you are not publishing the research results is not an option in most cases (item 6.3). In general, especially honours and masters students here seem to undersell the likelihood that their research will reach a public. The committee is committed to getting research out, so we’ll generally encourage people to say that they would like to publish their results.

#9. Recruitment, conflict of interest, and other power-related issues. In many anthropological projects, researchers are embedded in their communities in multiple ways; sometimes we conduct research among ‘our own people,’ sometimes we have other roles in the community while conducting research, sometimes we enter under the umbrellas of other groups already working in the field.

So the sections on conflicts of interest (section 4) and recruitment (section 5) are meant to sort out the possibility of coercion. Some populations (section 4) are likely to raise concerns for the ethics review board, especially students (if you’re their teacher) or employees (if you or the person recruiting them is their boss). In addition, the committee will be more careful with any of the populations indicated in 4.2: Aboriginal groups or Torres Strait Islanders, foreign populations, prisoners, asylum seekers, soldiers, the mentally ill or disabled. Although the committee WANTS responsible researchers to do work with these groups, we also want to make sure that they have the same protections as other participants, which may mean that the researcher has to go to a fair bit more trouble; for example, getting real informed consent from a population that is not free or does not feel free to refuse to participate is tricky.

The committee tends to worry about coercion, either implicit or explicit. Some members of the committee may even be overly concerned about coercion; my own experience in fieldwork is that people find MANY ways of refusing to participate in our research, not all of which are immediately obvious.

But the most chronic problem about coercion actually tends to come up with recruitment, especially when recruiting is being done through an organization or group representative. The committee doesn’t want people to think that getting access to resources (including therapy, treatment, or legal support) is contingent upon participating in research, so we tend to prefer a bit of an arm’s length relationship between contact people in organizations and the researcher. That is, we’d prefer that contact people give out information, but not actually sign up participants. We’d prefer that participants be asked to do something active in order to participate rather than force them to be active to opt out.

In general, in ethnography, if we explain our situation to the ethics committee, it’s not a problem. Approaching people directly, in most places, is not a problem, as long as you can explain to the committee that the participants have a relatively straightforward path to exercise in a right of refusal. For example, in an ethics application that I’m currently writing (and should be finishing instead of doing this), I’m proposing to interview a stonemason who I have contracted to work on the farm my wife and I have. Although he’s an ‘employee,’ and thus might be considered liable to coercion, in reality, our relative social positions in the community, the shortage of stonemasons, and the fact that he has a backhoe (infinitely valuable if you have a rural property) actually means that there’s no way I cold possible coerce him into participating if he didn’t want to. If I describe the social situation, it should be clear to the committee that this gentleman is virtually immune to any attempt to compel him to participate in research.

#10. Finally, the committee is committed to protecting the researcher and the integrity of research done under the auspices of Macquarie University, not just the subjects. Sometimes I think students (and even senior researchers) can lose track of this.

Applicants sometimes promise things that are unnecessarily onerous, especially those who are most dedicated to ethical behaviour. We sometimes suggest that a procedure for transcript review or a promise to give all of the subjects a copy of the resulting thesis is simply too difficult and disproportionate for the researcher. This may be hard for observers from outside our university to believe, but we think that sometimes the conditions imposed on research – even conditions imposed by the researcher – can be too burdensome. It is always better to deliver more than one promises than to make assurances in the field that cannot be followed up on. I know that, for any researcher who does believe strongly in ethical conduct, unfulfilled promises can weigh heavy on the mind and heart after ethnographic work. We learn so much from so many and amass so many debts, both personal and moral, that it can be very difficult to discharge even a fraction of them.

So although I would be the first to advocate for reciprocity and cooperation, I also will argue in committee meetings that some processes — for example, letting informants edit transcripts of notes and thus giving them veto power over research data or promising too many copies of any resulting work — might be either detrimental to the integrity of the research or aspirational targets rather than conditions of fieldwork.

Note: I’m going to post this as is because it threatens to grow longer and longer. I suspect that, if I get feedback, it may be critical, but I write this quickly in order to share some suggestions with students and colleagues here at Macquarie (and perhaps, by extension, elsewhere). With some time and greater thought, and perhaps with feedback from other observers, I may revisit these topics or even alter my suggestions.

The bottom line for me, however, is that the ethics review process can be a positive learning experience, one that allows applicants a chance to articulate their ideas, explain their plans, and receive feedback that helps to improve their projects. For review to work in this way requires good faith efforts on everyone’s part: applicants and reviewers alike. Whether or not we have achieved this in the past — and I’m sure that even our committee here has made decisions with which present committee members might take issue — we need to continue to work to make the process better.

I stand by my initial posting; that the answer to the problems of ethics review for ethnographers is not to avoid review, but to improve it, to educate committees, to educate applicants, to create a pool of shared knowledge about fieldwork techniques and ethics. My hope is that the process becomes steadily more sane, less intimidating, more educational, and better able to protect subjects, researchers, and the integrity of the research that we do as an academic community.

Greg Downey
23 August 2007