The case for ethics review in the social sciences: Drawing from practice at Queen Mary University of London

Abstract

This article responds directly to an article published in Research Ethics in 2011 where Schrag argued against ethics review for social science and humanities research. He argued that review committees offer solutions in search of a problem, impose silly restrictions and apply inappropriate principles. He suggests that review committees typically lack appropriate expertise and argued that the process harms the innocent. This article refutes these claims and offers a case study of the ethical review process at Queen Mary University of London (QMUL) to offer counter claims. The discussion highlights the way in which the QMUL process is sensitized to the challenges posed by social science and humanities research and is a process that, rather than focusing upon avoiding harm, emphasizes notions of care to both participants and researchers.

Introduction

Drawing upon his own experiences and those gathered via his blog, according to Schrag (2011), when ethics committees in universities around the world review social science and humanities research they are guilty of a number of charges relating to process, inputs and outputs. In terms of process, Schrag suggests that review committees offer solutions in search of a problem, impose silly restrictions and apply inappropriate principles. He suggests in terms of inputs to this process that review committees typically lack appropriate expertise, and in terms of outputs he claims that the process harms the innocent. He concludes by suggesting that even if it is the case that researchers in social science and humanities should be subject to ethics review then perhaps better options do exist in terms of: (i) who conducts the review; (ii) what questions are asked of researchers; and (iii) when the review takes place. In so doing he offers a number of alternative approaches which others have proposed would be more effective (Rasmussen, 2008; Tolich and Fitzgerald, 2006), and suggests that too often advocates assume or assert its benefits without providing examples.

In this response I will largely refute the charges that Schrag levels against ethics review committees by drawing upon my own and others’ experiences of the ethics review process at Queen Mary University of London (QMUL). In so doing I will firstly problematize some of the assumptions that are embedded in Schrag’s arguments, that suggest that he has fallen into the trap that many US researchers and commentators do, which is to claim to be non-partisan but, in practice, adopt a largely US-centric view of this issue. I will then demonstrate that although there may be great variation in practice, both in the UK and elsewhere, good practice does exist which supports both the ethical production of knowledge and the professional development of social science/humanities researchers. The practices which I outline here highlight that it is feasible to review social science and humanities research consistently, in a way that does not impose silly restrictions, and one which is commensurate with the degree of risk involved to all parties, when judged against the potential value of the research.

Embedded assumptions

The first assumption Schrag makes, supported by some examples of researcher’s experiences in other universities ¹ around the world, is that the somewhat draconian measures taken with respect to the ethical review of social science research in the US are applied equivalently elsewhere. This is not the case in the UK. Articles in Research Ethics have highlighted the variety of approaches that are adopted in UK universities (Stibbs, 2009), ranging from, in some cases what effectively amounts to a cursory, rubber stamp approach to review, to the application of what might be considered to be a ‘medical’ model to social science research. This is not, however, to concur with Schrag’s argument around the inability of ethics committees to be consistent in terms of decision-making, but is simply to point out that the examples he draws upon in making his assumption are highly selective. In practice a range of approaches are applied in the UK and elsewhere. I do not wish to suggest that this is necessarily good, but rather to temper Schrag’s assumption that all or the majority of social science research outside the US is subject largely to a medical model of ethics review and hence imposes silly restrictions. There will always be some evidence of silly practice to be found, but most researchers outside the US are not always, or even typically, subject to medically modelled review regimes. It is perfectly feasible therefore to suggest that institutional level reviews might exist that support the production of ethical social science research.

Secondly, Schrag (perhaps not surprisingly given his own background) implies that most social science/humanities research adopts an ethnographic approach. This is evident in that he draws largely upon ethnographic studies in supporting his main arguments. Although it is widely acknowledged that ethnography does pose some very distinctive challenges in terms of ethical review (Murphy and Dingwall, 2007), it is also important to recognize that in most universities social scientists apply an array of methodological approaches of which each may provide its own ethical challenges, which need to be specifically addressed. Although commentators often suggest that decentralization and peer review is the way forward here, often peers are blinded to the ethical implications of their own methods (Bok, 1999).Therefore offering researchers the opportunity to engage with a multi-disciplinary review of their research would seem to make sense.

Thirdly, Schrag assumes that everyone conducting social science/humanities research is well trained in methods and has a sophisticated understanding of what constitutes ethical research. In my experience as a doctoral supervisor in three UK universities, I have noted that even after formal research training, the majority of doctoral students often remain ambivalent about various general ethical principles which, regardless of the nature of the review process, constitute good, professional research practice. The same might also be said of some academics I have encountered. We often do not know what we do not know. When doctoral students embark upon their research design, whilst they may be clear about their ontological and epistemological position, simultaneously they often demonstrate their naivety regarding such ethical practicalities as to what might constitute informed consent or who might be considered vulnerable? By engaging in ethical review these matters become apparent prior to fieldwork commencing and can be jointly addressed by the researcher and those with experience and expertise in these matters.

Finally, Schrag assumes that the only harm that can be done is either to participants or the researcher. He fails to acknowledge that institutions are also at risk, in terms of both their duty of care to employees and students and their reputation if, for example, unethical research were conducted.

This, and the other assumptions which are embedded within Schrag’s arguments, suggest that despite gathering considerable data on his blog, and drawing upon numerous examples from the literature, his perspective on the ethics review process appears to be largely influenced by his own research background and experiences. By problematizing these assumptions I have highlighted that they largely echo and support what many other social scientists believe and previously articulated in a variety of articles (Dingwall, 2008; Ferdinand et al., 2007; Hammersley, 2009). However, in each case I have suggested that institutional level review is still perhaps the most appropriate approach to adopt.

I will address the charges that Schrag has made against ethics review and, in so doing, return to the assumptions that I have problematized here, by directly drawing upon my experience as a member of QMUL’s ethics committee and also the views of a number of students and social scientists who have participated in the QMUL ethical review process. I do not claim that the process and practices that have been developed at QMUL are typical in the UK. However, I aim to highlight that the approach that is used demonstrates how all manner of social science and humanities research can be effectively and efficiently reviewed when the emphasis throughout is placed on affording care rather than avoiding harm. Although on the face of it these terms might be considered equivalent, there are important semantic differences which I believe shape the ethical review process and practices at QMUL, which I return to in the Conclusion.

The QMUL ethical review process

The QM ethics committee is chaired by a highly experienced lay person, who is also a member of University Council. She has considerable professional standing and, because of her lay position and wide experience, she is able to offer a non-partisan view and clear guidance. She and the committee are also supported in their work by expert advice and support from a senior administrator. The remainder of the committee is made up of 13 academic members with considerable research experience, drawn from all faculties within the College, together with the President of the Student’s Union and the Vice Principal for Research. This composition ensures that the committee provides transparency, is able to offer multiple perspectives and can apply a broad, yet deep level of expertise, in terms of a collective understanding of research context and methods, when reviewing research applications.

Two processes are in place to handle what are referred to as low- and high-risk research. The majority of Master’s level projects and some academic research that is undertaken is low risk. This is determined quickly and efficiently by students and researchers completing an online questionnaire. Students provide a synopsis of their research project and answer a series of questions. If any of the questions are answered in the affirmative (or if the synopsis indicates that questions not answered in the affirmative should have been), then the project may not be low risk and the application is referred back to the researcher or the student and their supervisor for further consideration. In addition, the Administrator audits all low risk applications before giving formal ethics approval, which serves as a final check. This process effectively handled 203 applications in 2012, 46 percent of which were from social science and humanities students.

Where research is considered to be potentially high risk, then students and academics must submit a full application to the ethics committee. Each application, subject to timely submission, is firstly reviewed by the Administrator, who will highlight any fundamental concerns to the researcher, requesting any further information and clarification that it is likely the committee might require, prior to formal submission. This provides applicants with an opportunity to rectify anything before submission which might lead to an application being deferred or rejected outright, e.g. an incomplete submission, evidence of approval to access an organization etc. In addition, whilst students are expected to work with their supervisors in preparing their application, they, along with academics, can contact the Administrator at any time prior to submission requesting advice. The emphasis from the outset, then, is in offering support to all QMUL researchers.

Each application is then allocated to one member of the committee for an in-depth review, although all members are expected to read every application prior to the meeting. The reviewer will provide a verbal summary of the application and any concerns they may have. The reviewer is not necessarily from the same disciplinary background as the applicant. This, however, is not a drawback, and typically the reviewer is able to provide a complementary, rather than antagonistic, perspective on the ethics of the proposed research design. The committee debates the issues that have been raised by the reviewer and any further issues that other members may have identified. Typically, these are resolved quickly by drawing upon internal policy (often grounded in decisions which were taken on previous applications) and, importantly, on members’ collective perspectives, disciplinary expertise and experience. This approach therefore ensures, as far as possible, consistency of decision-making and it avoids imposing silly restrictions.

Researchers are invited to attend meetings and will be asked to respond to any outstanding issues or points of clarification the committee has raised at the meeting. The decision-making process is thus characterized by debate and, in some instances, negotiation both within the committee and with the applicant. The overarching aim is to actively engage with the life worlds of both potential participants and the researcher, whatever their disciplinary background and proposed methods. In so doing, then, rather than the committee adopting a liminal position, operating somewhere between these different life worlds, our goal within the QM research ethics committee is to adopt a duplex position (Parker, 2007). This is a position which recognizes that what needs to be negotiated is not only the ethics of the research but, importantly, its mode of knowledge production − something which Parker refers to as enfolding ethics with method. In this way care is afforded to both the researcher and participants. In order to provide an example of what this means in practice, we might consider the issue of offering anonymity to participants. Typically, applicants state that all participants will be anonymized, assuming this to be good ethical practice and something which the committee would expect in terms of research design. Although in the majority of cases we agree that full anonymization should occur, we also occasionally point out that in particular research contexts it might be more appropriate to offer participants a ‘voice’ and give participants the opportunity to make that decision for themselves, provided they are given a clear explanation of how their data are going to be used.

When research is considered to be very high risk, i.e. it is generally agreed across the committee that there is the potential for actual psychological or physical harm, either to participants and/or the researcher, or the university’s reputation may be put at risk, we engage in further processes which broaden the decision-making arena. For example, the QM Senior Executive will be asked to endorse high risk applications in order to establish whether they, with advice from the committee, believe the research to be of such potential value that the committee should work with the researchers to identify the means through which the research can proceed. This ensures that senior management are aware of potential reputational risks from the outset and any issues around the university’s duty of care. Two applications were considered very high risk in 2011. In both cases we decided to support the researchers and worked with them to set up an expert advisory panel to offer advice and guidance around the development of an ethical research design, which would minimize the potential for harm. In this way, we actively supported the researchers so that they did not enter potentially dangerous or ethically dubious territory. To put in place these measures, decisions on very high risk applications will then necessarily have to be deferred to a subsequent meeting, but the recommendations that are offered are all aimed at enabling the research to take place, while simultaneously ensuring the care of both the researchers and participants. By adopting the approaches described here that we refer to as high risk research, the vast majority of applications (94% in 2012) were either approved or conditionally approved at the meeting to which they were submitted. As we hold 11 meetings a year, ethics approval is then very rarely a cause for delay in commencing research at QMUL. We therefore do not look for problems as Schrag asserts, but in practice the committee is often faced with ethical problems around research design to which, as far as possible, we offer workable solutions, drawing upon our own, and occasionally others’, expertise.

I conducted short interviews with a number of academics and doctoral students in social sciences/humanities at QMUL in writing this response, and the overwhelming message was that by engaging in and with this process, they reflected extensively on both their research design and the roles/responsibilities they had as researchers. All of those with whom I discussed the process therefore viewed the experience as contributing to their professional development. In summary, Queen Mary’s research ethics committee does not impose silly restrictions. Rather, it provides solutions to genuine problems which may or may not have been identified by researchers and, as such, both contributes to researchers’ professional development and affords care to participants. In so doing perhaps the most important aspect of the process is that the committee adopts a duplex position, which combines sensitivity to both ethics and method.

Conclusion

Care is often described as the judicious avoidance of harm. Arguably, however, in focusing upon avoiding harm many review committees fail to give adequate consideration to the notion of care. I have discussed the way in which QMUL engages in ethics review and have argued that in terms of inputs, processes and outcomes we are able to refute the charges that Schrag levels at review committees. Schrag asked for examples of good practice, and what I have described here suggests that the process and practices, at least in one UK University, demonstrate sensitivity to the challenges posed by social science and humanities research. We aim to ensure that as far as possible all ethical aspects of the research have been considered and accommodated in the research design, so that:

• Participants are afforded care and respect.

• Researchers have the methodological processes in place to provide that care and have also taken adequate steps to care for themselves and the university.

By problematizing some of the assumptions that were embedded within Schrag’s article and highlighting that good practice does exist at an institutional level, rather than identifying better options as Schrag suggests, perhaps a more interesting question to further this debate is: What institutional or ethical barriers exist in the UK or elsewhere that prevent the approach described here from being adopted more widely?