Abstract
Current research ethics processes, based on the mantra of privacy and institutional protection, take a paternalistic approach to research participants that leaves them open to harm. Reflecting on our own research/consultancy as a case study to illustrate the current flaws, we explore our and our subjects’ experiences within the wider political context of institutional ethical rules and the Australian NHMRC guidelines. In doing so we argue for fundamental changes to the modern research ethics processes – a system that treats participants more as research collaborators rather than victims in waiting. A complete review of ethics processes is needed to empower participants and researchers to recognize the reality of the process as co-created and negotiated. This includes changes at the top level of research administration – a shift in ethics policies and procedures as well as greater education in ethics with commensurate trust for active researchers.
Introduction
Researching people constructs problematic power dynamics that can result in significant harm to participants, 1 researchers and the wider endeavour of accumulating knowledge. This has resulted in the implementation of ethical frameworks through which research must be conducted. Originally created to protect participants from harm, ethics processes are now ‘a risk management exercise at the behest of the host institution or funding body’ (Tolich and Fitzgerald, 2006: 72). These processes have fundamentally corrupted the research process and infantilized the subject and researcher - to such a degree that we must critically question all research done under the auspices of ‘informed’ consent.
It is fair that some form of regulation of research is required given the history of unethical studies (Babbie, 2004; Orb et al., 2001), which led to the development of ethical principles for research (Dingwall, 2006; Halse and Honey, 2007; Metro, 2014). However, like all regulatory regimes, activities outside these frameworks still occur, with a recent controversial example being the secret manipulation of Facebook feeds to assess emotional contagion (Fiske and Hauser, 2014; Harriman and Patel, 2014; Kramer, Guillory and Hancock, 2014).
Bio-ethics was the first area to be regulated (Ramcharan and Cutcliffe, 2001), being formalized in the Declaration of Helsinki (Gelling, 1999; World Medical Organisation, 1964) with social science research drawn into this medicalized framework later (Christians, 2000). These codes have generally covered four domains; informed consent, opposition to deception, safeguards to protect privacy and confidentiality and, ensuring accuracy in research results (Christians, 2000). Built within this framework, committees reviewing qualitative research still assume projects should emulate quantitative, medical-model research (Richardson and McMullan, 2007; Tolich and Fitzgerald, 2006).
Ethics processes were initially focused on the concrete harms of physical abuse and medical malpractice. Over time this expanded into social harms, with participant privacy and confidentiality becoming key (Baez, 2002; Giordano et al., 2007; Kaiser, 2009; Kelman, 1977). In general, ethics committees dictate that researchers are obliged to protect the identity of their research participants (Giordano et al., 2007; Ryen, 2004) and ensure that this protection is fully explained (Giordano et al., 2007; Rae and Sullivan, 2003). Research standards assume that participants will stay anonymous during the research process. These concerns are based on a value-judgement about the role of privacy and secrecy in our society. Social research is seen as inherently intrusive (Kelman, 1977) with participants sought by researchers for their own purpose (Guillemin and Gillam, 2004: 271).
However, qualitative researchers have moved beyond these narrow concerns, with a focus on the problem of power in the researcher-participant relationship (Lincoln, 2009). This new ethical debate ranges from discussion on the sharing of interview transcripts with participants (Forbat and Henderson, 2005) to the development of collaborative interview and research processes (Bosworth et al., 2005; Bresler, 1995; McTaggart, 1997; Munhall, 1988; Salmon, 2007). These methods shift research away from a position where researchers exercise power and instead frame research as the joint development of knowledge between participants and researchers (Acker et al., 1991; Lather, 1988; McTaggart, 1997; Salmon, 2007). This makes the research agenda about social empowerment first, knowledge-gaining second. Research should directly empower its participants, with ethical frameworks ideally designed to facilitate this process. Yet, despite decades of argument on these topics, we still operate under a framework that fails to empower.
In Australia, research is governed by the Australian Code for the Responsible Conduct of Ethics (National Health and Medical Research Council, 2007a), jointly developed by the National Health and Medical Research Council (NHMRC), The Australian Research Council and Universities Australia. The code provides the key standards for research institutions and researchers alike (National Health and Medical Research Council, 2007a), outlining responsibilities and mandating that research institutions ‘provide an appropriate research governance framework through which research is assessed for quality, safety, privacy, risk management, financial management and ethical acceptability’ (National Health and Medical Research Council, 2007a: 1.3). The National Statement on Ethical Conduct in Human Research (The National Statement) outlines the establishment of Human Research Ethics Committees (HRECS) within major research institutions, mandating that ‘each member of an ethical review body is responsible for deciding whether, in his or her judgement, a proposal submitted to the review body meets the requirements of this National Statement and is ethically acceptable.’ (National Health and Medical Research Council, 2007b: 74).
HRECs operate at a rather narrow level, being required to assess research proposals based on two criteria set out within the National Statement: the risks of the research in comparison to the benefits, and ensuring that measures are taken to ensure participants are able to make informed decisions about participating in the research (National Health and Medical Research Council, 2007b). These two areas of assessment are based around the premise of harm. Harm operates at a macro and micro level, first the broader question of intervention in circumstances of exploitation (i.e. observation of harm) and second, ensuring that no harm is done via the research (Bresler, 1995; Orb et al., 2001).
Ethics processes were developed because of the need to regulate researcher conduct, however, the risk averse, legalistic and harm focused models we now have do not cater for the needs of qualitative research. We critically reflect on our own experiences with a problematic research/consultancy engagement that was ethically approved, yet did not empower the research subjects. Situating our argument in the more radical and critical ethics literature we show how current ethics processes need to be overhauled. We add to the literature on ethics processes by reflecting on an Australian experience which allows us to critique the NHMRC guidelines specifically. We also reflect on the grey area of research/consultancy work and the pressure to take research funding from any source in the current higher education funding climate and how this type of research is poorly catered for under the guidelines.
Case study
A number of ethical issues became apparent in a research project we conducted. These issues were specifically raised with us by Martin Tolich at a TASA conference (Connor et al., 2012). Tolich is noted for opening debate on the issue of internal confidentiality, which he describes (2004: 101) as ‘the ability for research subjects involved in the study to identify each other in the final publication of the research’. An often unspoken part of the confidentiality puzzle, Tolich believes that internal confidentiality needs to be brought into a broader debate, with ethics processes making it a part of the confidentiality process as it widens the notion of vulnerability (Tolich, 2004). Internal confidentiality (or deductive disclosure, Sieber, 1992) is of particular concern for the workplace, where criticism of superiors, leadership techniques, workmates or the work of the organisation from participants, can lead to reprisal (Kelman, 1977). Venkatesh (2008) offers a sobering insight into the risks of communicating information to senior leadership that junior members of an organisation thought was offered ‘in-confidence’, when he revealed information to the gang leader.
Our research focused on a change management process of a Federal Agency in the Australian Government. Through a series of research agreements we conducted an analysis of an IT heavy program within the organisation. The Agency was trying to fundamentally change how it operated, shifting from a paper-based system to a computerized one. We were approached to review the Program and provide consultant-like reports on its progress and effectiveness. As part of the contract we negotiated the right to use the data we collected for research purposes and ultimately for publication. The studies were conducted via background discussions, focus groups, workshops, in-depth interviews, and, analysis of internal and publicly available documents covering the Program. The research program was supported by the highest levels within the Agency, giving the research team unprecedented access to subjects.
The first research cycle consisted of six workshops each taking two hours. These were conducted with the various players/teams in the Program and canvassed individuals’ views from the project teams involved. Twenty in-depth interviews, each taking approximately 90 minutes, were also conducted. The second cycle, approximately 18 months later, consisted of 15 follow up interviews. All interviews and workshops were taped, transcribed and coded based on key themes derived from the literature and those that emerged during the research. Two detailed reports as well as verbal briefings were delivered to the research partner as part of the agreement. These were considered the main output of the research process, and included a range of recommendations for senior management for the following tranches of the change management program and for the organisation more widely. Second, the authors used research data from the process to develop a range of academic papers analysing the process.
As we intended to use the data we collected for research (as well as consultant reports) we applied for, and received, ethics clearance from the UNSW HREC. With the Agency requiring us to begin work within weeks of first approaching us this application was filed and approved quickly. Ethics procedures occurred simultaneously with the drafting of contracts with the Agency and the development of research structures and interview questions. This fast timeline ensured we as researchers did not have appropriate time to adequately consider many of the ethical questions associated with the project. This fact was compounded by financial requirements to take on the project. In a time where universities and researchers are increasingly compelled to seek outside funding in order to conduct research we were unable to reject this project.
As the research was conducted, we took active steps to ensure the confidentiality of subjects, with assurances of confidentiality for all subjects (in writing and verbally). We used the standard UNSW Participant Information Statement and Consent (PISC) forms and terminology. Subjects were recruited to the study by the Agency itself, conducted on site during working hours - which implied a level of compulsion. While we made it clear that participation was voluntary, no subject refused – raising questions as to how voluntary the research was.
Different levels of confidentiality were used depending on the research output. In reports presented to senior management all names and position titles were removed from the documents, with quotes only attributed to ‘an individual’. In research papers however, we provided more detail on subjects’ roles in the organisation. This was done to provide greater context, which we believe was required for the proper reporting of the data to a wider audience. Both of these approaches were in line with agreements signed by subjects.
Discussion
Our case study highlights a significant number of problems both with the standard nature of ethics practices, and in particular the implementation of these practices within a researcher - consultant/audit framework. On reflection, we have come to the conclusion our research had a number of ethical problems which we failed to address. In our defence the current guidelines make it nearly impossible to research ethically, thus we, like so many other researchers, accepted institutional approval as a bureaucratic requirement to be met as a tick-a-box exercise.
First, we did not gain true informed consent of our subjects. As all subjects were recruited through The Agency, with formal requests for participation being made by Agency management, it is reasonable to believe subjects understood participation to be a requirement of their employment. There is a distinct power differential between Agency management and staff, which was exploited during the recruitment phase. Even though we offered subjects the opportunity to withdraw at any time, we believe concerns about retribution within the workplace could have stopped subjects from doing so. Miller and Bell (2002) offer another case study of possible work-place coercion and they chose to change their consent procedures to account for this, we should have considered this ourselves.
Second, our research put subjects at the risk of being internally exposed. While individual names and position titles were removed from research documents, it is plausible that internal audiences may have been able to triangulate documents in order to identify subjects. This presents two risks. One, senior management may have been able to identify informants. This is particularly relevant as our original consultant reports were only provided to senior management, meaning Agency staff did not necessarily have access to the details, nor the recommendations provided. Two, the externally published papers/presentations may allow for triangulation of who said what and allow other staff to identify subjects. This had the potential to extend any form of backlash to informants to their colleagues as well as their employer.
Many subjects were deeply critical of the Program and, in particular, the role of senior management. The identification of individuals could lead vengeful members of senior management to punish these subjects – including the potential removal of responsibilities, the denial of promotions, or isolation from colleagues and work experiences, amongst others (Kelman, 1977). This can also be extended into other parts of the workplace, with many subjects criticising the work and practices of fellow colleagues or work teams. Indeed, a key theme of the data was the denial of responsibility and blame shifting engaged with by many subjects - there was always someone else in the organisation to blame for the Program failures/delays. These risks are particularly relevant given the assumption subjects would have had that assurances of confidentiality would have protected them. The lay understanding of what confidentiality means can be very different to the reality, with subjects not interpreting it the way ethics guidelines assume (Sankar et al., 2003; Swain, 2006; Wiles et al., 2008).
Finally, we believe we failed to properly empower and inform our subjects during and after the research process. Research findings were given solely to Agency management (in the form of an audit report) or published in hard-to-access academic journals, leaving subjects completely absent in the decision-making process that followed. In her research with Aboriginal women in Canada, Amy Salmon noted a perception of researchers as being people ‘who collect information and “data” from Aboriginal people but “don’t give anything back”’ (Salmon, 2007: 983). It is a common obligation for researchers to return something to subjects. In our work it was up to senior management as to how and when the findings were to be distributed internally (if at all) – separating subjects from the data collected. Subjects were passive actors, with decisions about outcomes completely out of their hands. Whilst management would have been able to make changes following the recommendations, the workforce did not have that opportunity. This, we believe, goes against many of the values embodied in social qualitative research.
Exploiting ethical gaps
We did not engage with our subjects as well as we could have, with the potential for our research/consultant methodology to have negative consequences. However, when asked the question of ‘what did we do to fix this?’ the answer is nothing, as technically we did nothing wrong. Our research received ethics approval and we ticked all of the appropriate boxes in terms of national regulation and University policy.
This highlights major problems with standard ethics procedures, much of which was exacerbated by the nature of our research. As researchers hired by the Agency on a consultancy our research fell in what we called the researcher-consultant ethics gap.
Engaging in an audit, we applied for ethics approval to use the data for academic purposes. This placed us in a problematic position of attempting to engage in consultant-based work, whilst maintaining ethics procedures normally designed for academic social research. It is here where many of our problems arose. We were hired by the Agency on a very short turnaround requiring us to rush through approvals (days, not weeks). We did not have the time to properly think through the ethical conundrums within the project, missing many of the issues noted above. Second our auditing contract created a number of the ethical problems we encountered – in particular the recruitment of participants and the reporting of data to Agency management. This left us with challenges we were simply unable to resolve. Despite this, we were still able to receive ethical approval.
It is important to note the motivation for undertaking this research. Performance expectations of researchers in modern universities are focused on two key outputs: research income and quality of academic papers produced. Given the expectation of our university to bring in research funding this project was considered too good to pass up. This provides some indication of our motivation, and in particular highlights the conflict we faced when dealing with ethical issues. We note this problem not as an excuse but as explanation for the pressure we all feel.
Finding the gaps
Given the origins of ethics approval frameworks, it is not surprising that qualitative research that straddles audit/consultant work and research is not well covered in the guidelines. Originally written for the medical context, Australia’s first ethical code was developed in the early 1990s by the National Health and Medical Research Council, then without consultation or negotiation extended to include social research (Van den Hoonaard, 2011). This was further formalized in The National Statement produced in 1999 and updated in 2007. The National Statement covers all research including humans, whether quantitative or qualitative, medical or social, providing a base set of guidelines for research on humans.
The National Statement provides a brief discussion on qualitative research in Section Three: Ethical Considerations Specific to Research Methods or Fields. This section defines qualitative research as that which ‘can involve the studied use and collection of a variety of empirical materials such as case studies, personal experience, life stories, interviews, observations, and cultural texts’ (National Health and Medical Research Council, 2007b: 23). In doing so, it outlines some potential ways in which qualitative researchers treat ethics differently from their biomedical colleagues. These include the use of verbal consent, variations of subject recruitment and sampling sizes, discussions on the role of researcher-participant relationships and the potential for a researcher to be a therapist for participants.
The National Statement provides a very broad brush approach to ethical guidelines, outlining key principles for those working with humans. This is described in the section on Ethical Values and Principles, which says:
These ethical guidelines are not simply a set of rules. Their application should not be mechanical. It always requires, from each individual, deliberation on the values and principles, exercise of judgement, and an appreciation of context. (National Health and Medical Research Council, 2007b: 11)
Despite this, it is the general experience of social researchers to be forced into a one-size-fits-all legalistic approach. We identified three problems with our research: consent, confidentiality, and the researcher-participant relationship. In each of these areas standard ethical procedures allowed for little deviance in practice. The National Statement mandates that all research that is declared to be above low risk must receive ethics approval from a HREC, which produces a standardized measure of ethics, based largely on expectations of quantitative medical research (Bamber and Sappey, 2007). There is an expectation of confidentiality for participants, a practice of informed consent which occurs at the start of research via formalised consent forms. Ethics procedures force researchers into a series of templates, creating what Van Den Hoonaard (2011: 45) describes as a ‘blizzard of paperwork’. This bureaucratic process codifies a researcher-subject relationship, discourages difference, and restricts the involvement of participants in the negotiation of ethical procedures.
These realities are the result of ethics being managed with a risk-based (Bamber and Sappey, 2007), managerialist approach. Bell and Bryman (2007: 66, citing Furedi, 2002) argue ‘universities have developed an obsession with research ethics driven by risk assessment and a fear of litigation that leads them to act more conservatively than the law requires, being more concerned with the reputation of the university than the ideals of research’. Furedi (2002) notes that
review bodies are not in the business of discussing the finer points of ethics. They are concerned with issues such as risk, litigation, the reputation of a university and the management of research. In practice, they are not so much an ethics committee as bureaucratic gatekeepers who use ethics as a managerial ideology.
As Iphofen (2011) notes, institutional integrity is the aim of the university, this means its focus is reputation first, ethics second.
This approach limits our ability to do good research, in particular, research that involves, empowers and challenges – as that is seen as risky. Yet qualitative social research is often significantly less risky than its quantitative medical sibling, as Dingwall (2006: 51) notes: ‘risks to human subjects are not comparable [with medical research] and the power relationship between researcher and researched is so different as to render prior scrutiny irrelevant and inappropriate’. Largely concerned with the legal, financial and reputational risks of ethical breaches, universities maintained the same ethical procedures, even though they were not suitable for social research.
Managerial ideology places participants in the role of subject, to be protected from the potential harms of research. This has a major impact on practice, with qualitative researchers seeing the roles of researcher and participant as mutually exclusive: researchers define the project and contribute all of the thinking towards the end result, while participants are subjects whose actions and words are to be studied (Karnieli-Miller, 2009). This is exemplified by how ethics approvals are still sought before any engagement with participants. Researchers define what is ethical, with participants having limited say.
In treating participants in this manner ethics processes act paternalistically (Giordano et al., 2007) and assume participants are unable to manage the impacts of expressing their voice. This goes against much of the ethos of qualitative research, which should be the co-creation of knowledge (Ebbs, 1996). The interview in particular allows participants and researchers to engage in intimate, therapeutic and collaborative processes (Birch and Miller, 2000; Karnieli-Miller, 2009; Salmon, 2007; Tang, 2002), ones that are enhanced by techniques such as sharing interview transcripts with participants (Forbat and Henderson, 2005). This not only helps researchers gain and build on the knowledge required for their work, but allows participants to take ownership of their knowledge, so they can ‘come to terms with historical, social and cultural contexts of their communities and their position within those contexts’ (Ebbs, 1996: 219). The process of assumed confidentiality, which is still the basis of HRECs, disengages participants from the process by silencing their voices (Aldred, 2008). Our research was only shared with Agency management and not with employees – i.e. the majority of research subjects. Journal articles, often placed behind a paywall and written in confusing, jargonistic language, are also largely inaccessible. Participants are removed from the process of knowledge co-creation and have little to no access to the knowledge when it is created. These problems are well documented within the research community, with some researchers now deliberately disobeying ethical guidelines (Katz, 2006; Metro, 2014). These ethical outlaws are removing themselves from ethics processes, engaging in what they believe to be far more ethical behaviour than is prescribed in formal processes. In the Australian context, you would be very brave to risk the opprobrium of a University Ethics Committee and misconduct investigation – with the likelihood of losing your employment.
This reality has, potentially unintentionally, been reinforced by Government regulation and The National Statement. With the National Statement drafted with a focus on scientific, quantitative research, with social qualitative research being brought into the fold with little consultation from social researchers, qualitative research is automatically side-lined as something different or unique, and in turn something that can be ignored. The National Statement therefore plays a role in maintaining what Van Den Hoonaard (2011: 48) describes as the ‘hegemony of ethics committees’.This is the process in which the ethics application process ‘(even for minimal risk, non-treatment studies) has become “normal,” even when many parts of it are considered unreasonable (and even impossible) for some kinds of studies’ (Hamilton, 2002: 186–187). The risk-adverse, harm-elimination, scientific approach to ethics have become codified within the qualitative experience via Government regulation and most importantly the practice of increasingly invasive HRECs. This is occurring despite the common concerns of researchers regarding the failures of the system.
Closing these gaps – creating a qualitatively friendly ethics system
The current system gave us little choice in how to approach our research, especially given the time constraints and consultancy/research mix. We were forced into the blizzard of paperwork, filling out application forms in a rote manner like we have many times before. This turned the process into a tick-a-box exercise with little reflection. This discouraged both us as researchers and our HREC from actively thinking about the ethical issues facing our project and how we might resolve those. We have identified four initial areas of change to make the process more effective; a new National Statement, a rethink of the way HRECs operate, compulsory ethical research training for social researchers and researcher reflexivity.
First, there must be a new National Statement which focuses on qualitative research. An example of this can be found in the Canadian ethics processes. The Canadian equivalent of The National Statement is titled ‘Ethical Conduct for Research Involving Humans’ (Canadian Institutes of Health Research, Natural Sciences and Engineering Research Council of Canada and The Social Sciences and Humanities Research Council of Canada, 2014). A more extensive document than The National Statement, ‘Ethical Conduct for Research Involving Humans’ involves far greater discussion on qualitative research. While still having problems (Van Den Hoonaard, 2011) this section at least aims to deal more deeply with the complex issues facing qualitative research, including the timing of ethics reviews, differing forms of consent, the potential process of disclosing participant identities, the role of emergent design and the processes involved in observational research. The document deals effectively with the issues of qualitative research rather than seeing it as a simple side-line to quantitative approaches.
While the National Statement has glacially shifted on these issues since inception, these changes must go further, acknowledging the significant differences that exist between social and bio-medical research and adjusting the rules appropriately. This, we believe, should result in the development of a qualitative-specific national statement, which is developed in full consultation with social scientists and incorporates discussions on the issues we have raised. Importantly, as occurs in the Canadian situation (Van Den Hoonaard, 2011), this statement should act as policy not regulation, allowing for universities to vary procedures as issues arise.
Second, the way HRECs operate must change. Instead of the current system based on researchers filling out forms that specify the exact details required for research to be approved before undertaking any research, we should shift to an on-going, iterative system. One model is Tolich and Fitzgerald’s (2006), who argue for a process in which interviews are conducted between ethics committees and researchers. These interviews would consist of three basic questions, (1) what is the research project about; (2) what ethical issues does the researcher believe are raised by this project; and (3) how does the researcher plan to address these problems (Tolich and Fitzgerald, 2006). This gives researchers the opportunity to explain their research in an open-ended discussion with ethics committees, giving them the space both to think about and explain the ethical issues (Tolich and Fitzgerald, 2006). Instead of defining right and wrong ethical outcomes, this sort of process would allow for discussion between HRECs and researchers, with the two coming together to find appropriate solutions. This has the potential to be a fruitful peer supportive and collaborative process. For many institutions this would be a significant leap. In current systems, researchers are seen largely as untrustworthy actors, who must be monitored to ensure they do not breach ethical codes (Bamber and Sappey, 2007).
The composition of HRECs must be adapted to ensure they are run and led by fellow qualitative researchers. The current member composition requirements of HRECs in The National Statement are problematic as there is insufficient expertise (Bamber and Sappey, 2007) to administer a collaborative process (National Health and Medical Research Council, 2007b: 71–72). Schrag (2011a: 124) notes the problems this can cause, citing Atran who proposed research on international terrorism, which ‘was blocked by a board chaired by an expert in “the effects of hydrophobic and hydrophilic glass coatings, window tinting, and defrosters/defoggers on visual performance and driving behavior”’. Part of this is because HRECs operate in a largely centralized manner, meaning committee members are required to devote significant time and energy to the process. De-centralising HRECs and making them more like the peer-review process may ameliorate this, allowing experts to devote a small proportion of their time to participation in committees.
This will allow HRECs to apply different principles depending on the research conducted and assess harm depending on social good. For example, how does a researcher deal with this principle when researching police violence towards minorities (Moskos, 2008) or employer abuse of their employees (Bamber and Sappey, 2007)? Exposing these injustices will clearly lead to harm to participants (police officers, employers etc.), but there is also a greater good. Current HREC procedures would neither allow the observation of these practices (as it would be required to occur covertly) nor the disclosure of such practices, resulting in social scientists becoming servants of power (Bamber and Sappey, 2007). Discipline-specific HRECs may be able to deal with these questions more thoroughly, allowing researchers to find more appropriate solutions to these problems.
Further, we as researchers must take more responsibility for our qualitative methodologies. Researchers often enter their field with limited training in ethics, instead basing their understanding on the paperwork provided by HRECs and corridor scuttlebutt. Researchers require tools to build strong, trust-based relationships with participants, which needs to be part of researcher training. This will explain how to provide support for participants who have suffered emotional harm due to the research process (Birch and Miller, 2000), ensure participants have the proper capacity to provide informed consent and, most importantly, make research a valuable and empowering process for participants as well as researchers. Such a program must be compulsory for researchers, ensuring all who engage in potentially higher risk human research have the tools required to conduct ethical research. There are of course risks to such an approach. Schrag (2011b) has, rightfully, criticised the CITI training program, which Columbia University requires researchers to complete every three years. Schrag (2011b) quotes from a statement from researcher Ronald Bayer, who points to the prescriptive nature of the CITI program; one which defines set ethical answers and in turn doesn’t allow for any form of debate or discussion. Bayer describes taking the course as ‘the most insulting experience’, equating it to the process of children learning the Pledge of Allegiance by rote, whilst never actually understanding the meaning of the words. Conversely, the researcher training at Macquarie University treats ethics as an open, debatable issue, opting for open ended rather than multiple choice questions. Schrag (2011d) also points to an ethics program taught to PhD students in Princeton, one which is ‘discipline-specific, based on scholarly accounts of real controversies, and taught in person by expert faculty’. Taught as a workshop running over two days this program avoids many of the pitfalls of online courses by allowing students to engage in in-depth ethical discussions, developing ethical frameworks rather than simply promoting templates based on yes/no answers.
Using these examples we can see the principles that could be used to develop effective training programs. First, programs should ideally be completed in person, allowing for open debate, discussion and questions. Given the substantial resources universities place into ethics administration, as well as the time most academics dedicate to the blizzard of paperwork, the implementation of in-person programs is entirely reasonable, particularly if they replace other administrative functions. In addition to this, processes should not be designed with the idea that students/researchers must pass a course to have succeeded. This encourages the development of tests with correct answers, which is not applicable to most qualitative ethical conundrums. Instead programs should be based around giving researchers a framework from which to work, allowing them to explore ethical issues in depth on their own when engaging in their research. This would be a competency based program, not a curriculum based assessment package. Finally, any training program must be discipline specific, including the use of relevant examples. As Schrag points out (2011c) many of even the best training programs still use biomedical examples for qualitative research, even though these are largely irrelevant. Programs should therefore be aimed at specific disciplines, even within social research, allowing researchers to explore issues that are specific to their field.
In line with better education, we as researchers must become more reflective (Archer, 1995; Guillemin and Gillam, 2004) and engage slow scholarship (Harland, 2016) when we can. At the time, we as the research team had little opportunity to reflect on our processes, with the temporal criticality of starting the work over-riding our ability to think through the research. Now upon reflection, we should have more effectively communicated the confidentiality process to subjects and the Agency, in particular the risks of identification. We should have considered a multi-loop process of collection, cursory analysis, subject feedback, further analysis and so-on. This would have enabled a better sense of ownership and voice from subjects. However, the pace of work and the pressure to start precluded us from being reflective.
These four shifts; the development of a new National Statement, the reshaping of HRECs, the introduction of compulsory ethics training for qualitative research and reflexivity are essential for renegotiating how we re-fashion our currently broken ethics system. The current system does not meet the needs of either researchers or participants, making research frustrating for researchers and disempowering for participants. Importantly these systems are fundamentally failing to do what they have been set up to do: ensuring research is conducted as ethically as possible. We believe these changes can make the first important steps to addressing these issues.
Conclusion
Our case study highlighted serious problems with research ethics processes. As researchers we engaged in a range of potentially problematic practices that could have resulted in harm towards our subjects, yet, we did so with full approval from our relevant ethics panel. While some might argue this indicates a need for greater controls on the process of confidentiality, we argue this is a reflection of systemic failures in how we treat ethics in qualitative research. Current processes, based on the primacy of privacy, take a paternalistic approach to research participants. This is combined with a risk-averse, legalistic and fixed bureaucratic system that cannot allow for the nuance of qualitative research.
As Babbie notes (2004: 17): ‘While it is important to have codes of ethics, simply following established canons of ethics can degenerate into ritualism, which soon begs for clever shortcuts and other innovations that do not technically violate the canons’. A complete overhaul of the ethics process is required, one that fosters relationships between researchers and their participants, giving participants greater say over practices. At the most basic level it would require more in depth training for qualitative researchers to ensure they have the correct tools to engage participants. This can only occur via significant changes to the formal policies and procedures both of the Australian Government and then of individual universities and their ethics processes. These approaches will help deal with the many ethical problems researchers are facing, whilst ensuring the integrity of qualitative research.
Footnotes
Funding
The research reported on was funded by the Australian Governement Agency studied. However, the research team owns all the data and can publish as they wish – there is no conflict of interest.