Ethics in Online Research; Evaluating the ESRC Framework for Research Ethics Categorisation of Risk
by Kate Orton-Johnson
University of Edinburgh
Sociological Research Online, 15 (4) 13
<http://www.socresonline.org.uk/15/4/13.html>
10.5153/sro.2261
Received: 1 Nov 2010 Accepted: 11 Oct 2010 Published: 30 Nov 2010
Introduction
1.1 The ESRC Framework for Research Ethics (FRE) categorises research involving respondents through the internet as by definition research that would normally be considered to involve more than a minimal ethical risk. The document sketches out its logic for this categorisation by suggesting that issues of privacy, informed consent and uncertain participant identity pose new and unfamiliar ethical questions. While some aspects of internet research may be particularly ethically vulnerable, the FRE fails to engage with academic debates across the social sciences and humanities that have focused on the challenges and ambiguities of navigating ethically appropriate research online. In its assumption that all forms of internet research are inherently problematic, the FRE neglects the methodological and disciplinary breadth of web-based enquiry and, in doing so, threatens to tar a number of research settings and tools with too strict an ethical brush. Research in online spaces often focuses on substantive areas that have long been a staple of social enquiry and internet research faces many of the same methodological and ethical problems as ‘offline’ research. Risk categories identified by the FRE include research involving vulnerable groups, research involving sensitive topics, research involving groups where permission of a gatekeepers may be required, research involving some measure of deception and research undertaken outside of the UK. These are all areas where internet research, in common with research in more traditional settings, may be potentially vulnerable to ethical breaches. The logic of a priori labeling all internet research as potentially problematic, however, is unclear and unjustified and fails to recognise the fluid boundaries of online research methods and the considerable body of interdisciplinary work that has debated the complex terrain of ethical research online.1.2 In this response to the FRE, I will draw on the Association of Internet Researchers (AoiR) recommendations for ethical decision-making in a discussion of pivotal ethical concerns around informed consent and privacy in online methods. In doing so, I shall emphasise the need for a more detailed and considered approach to managing ethical conduct for research using internet technologies.
“Through the Internet”?
2.1 The problematic starting point in the FRE is the implication that ‘the internet’ is some kind of homogeneous monolith. The FRE’s categories of research at risk lists ‘research involving respondents through the internet’ (2010:1.2.3 p9) with a broader description in the Frequently Asked Questions of ‘internet research’ and ‘other research using new technologies’ (2010:p32). This extremely limited definition, and the treatment of internet research as a single category, fails to distinguish between research that treats online spaces and interactions as an object of social research, and research that uses technology as a methodological tool. The tensions of understanding internet technologies both as texts and as cultural spaces have long dominated methodological discussions about online research, but these important distinctions are not accounted for in the ESRC’s FRE. Internet research, as an umbrella term, draws on a range of methodological techniques, spanning diverse substantive areas and disciplinary boundaries. Immersive online ethnographic research or virtual ethnographies may have little ethical common ground with research that seeks to transfer traditionally offline tools, such as interviews, focus groups or surveys, into online fora. Similarly, empirical studies looking at language and discourse in computer-mediated communication (CMC) will have very different ethical concerns from research using web sphere analysis to study patterns and relations of cultural production. Grouping together, for instance, rich ethnographic accounts of internet use, with research that employs the global reach of the internet as a cost effective survey tool, seems a nonsensical basis on which to make judgements about ethical risks and seems at odds with the recognition in the FRE of the evolving nature of the field.2.2 Rather confusingly what is grouped together by the FRE is internet research and the use of visual data. The internet category of risk includes online research where visual images are used (2010: pg.9) and the ethics checklist appears to conflate internet research with visual methods by posing the question, ‘Will the research involve respondents to the internet or other visual/vocal methods where respondents may be identified?’ (2010: pg.34). It is not clear what visual or vocal methods are being referred to, but if participant identification is the ethical issue at stake then this demonstrates an ignorance of the strategies employed by internet researchers to protect participants’ anonymity and of the debates around the nature of identity representations and use of pseudonyms online.
2.3 In its treatment of internet research as a homogeneous category, the FRE departs from contemporary conceptualisations of technology and culture that emphasise the ways in which technologies have been normalised and incorporated into everyday life. The increasingly diffuse boundaries of technology and culture mean that ‘offline’ and ‘online’ have become less meaningful distinctions. Methodologically, treating online and offline spheres as separate spaces of enquiry limits our understanding of how technologically-mediated communications interact with, impinge on and transform social practices across a range of substantive issues. The importance therefore of conceptualising the social world as a continuum of mediated technological and non-technological spaces and interactions is key. This recognition is not simply a conceptual point, for it also raises exciting methodological questions about how researchers design and manage online field sites. The FRE would make this kind of flexible and adaptive approach to research design problematic to justify, and also in failing to recognise this continuum it threatens to act as a barrier to a critical methodological and substantive understanding of technoculture.
2.4 Stepping away from debates about online spaces as qualitative field sites, the ill-defined terminology of ‘through the internet’ has implications for a range of methodological tools that internet technologies have made available to the social researcher. Research designs may employ online sampling techniques or use quick and cheap online questionnaires to scope the field. Research has the opportunity to be multi-sited and multi-modal, and an ethical bias against virtual or online methods overlooks the importance and the implications of moving relationships with respondents between the online and offline fields in an increasingly mediated social world.
‘Ethical Decision-Making and Internet Research’ AoiR
3.1 The AoiR ethics committee produced the document ‘Ethical decision making and internet research’ in 2002 following extended multi-disciplinary, international debate around questions of ethical conduct in online research. This document provides researchers with a series of questions to guide ethical decision-making and, importantly, emphasises the need to reflect the methodological pluralism of internet research with an ethically pluralist set of guidelines:‘The issues raised by Internet research are ethical problems precisely because they evoke more than one ethically defensible response to a specific dilemma or problem. Ambiguity, uncertainty, and disagreement are inevitable…we hope that our work will help researchers…develop ethical responses to their specific dilemmas in Internet Research, especially as the distinctive characteristics of Internet Research and its highly interdisciplinary character make it difficult to apply extant guidelines to these new contexts.’ (AoiR ethics working committee[1])
3.2 The AoiR guidelines are markedly different from the absolutist stance of the FRE, falling between the two ethical perspectives allowed by traditional ethical models: ethical absolutism and ethical relativism. Its recommendations instead advocate an ethical stance based on the developing ‘moral career’ of the research involved, providing an ethical strategy that acknowledges unforeseen problems and takes into account the rapidly developing nature of new technological research environments and methods.
3.3 With their emphasis on ethical pluralism and cross-cultural awareness, and with their recognition that ethical judgements in internet research cannot be reduced to general rules applicable across all settings, the AoiR guidelines depart significantly from the FRE. The specific questions posed by the AoiR document provide a useful starting point for thinking about how Internet research has been framed and over-simplified by the ESRC, particularly around pivotal issues of informed consent and tensions around definitions of what is public and private online, identified in both documents as particularly ethically problematic for online research.
Informed Consent
4.1 Informed consent presents a particular challenge to ethical research in the online field and questions of what informed consent means and entails in online research are posed by both the ESRC FRE and the AoiR recommendations. Issues around negotiating informed consent are not unique to internet research and in many social settings the ethically responsible researcher faces the challenge of maintaining informed consent without jeopardising the validity of their data. However, the perceived ease with which covert research can be carried out online and the potential for researchers and participants to assume anonymous or pseudonymous identities has pushed the issue of informed consent to the forefront of debates about online ethical conduct. While the ideal of informed consent can be problematic in all forms of social enquiry, the transient and ephemeral nature of many online environments, often combined with large, fluctuating, unknown and disembodied populations, can make this ideal particularly problematic. For some online research settings informed consent poses an ongoing ethical challenge which demands reflexive attention to the role and identity of the researcher and to relationships with research participants and other users of the online space(s). However, other online research settings will face exactly the same ethical issues as their offline counterparts and here issues of context, identified in the AoiR recommendations but missing in the FRE treatment of internet research, are key.4.2 Informed consent rests on the principle that participants have information about research procedures, purposes and risks, that they understand what they are consenting to, and that their consent is given voluntarily. Online, verification of the competency, comprehension and particularly the age of potential participants has prompted some of the more sensationalist concerns around informed consent. However, as Walther argues, the extent to which identity deception actually occurs in internet research is unknown:
‘the degree to which these misrepresentations take place across internet research contexts is (a) an empirical phenomena as yet little explored, (b) probably highly inflated in public perception; (c) questionably linked to the motive to present dishonest responses to research questions; and… implies an ahistorical and naďve view of alternative research methods which have dealt with the same problem for many years.’ (Walther 2002: pg.211)
4.3 Meaningful informed consent involves interactions and relationships that are not easily reduced to a consent form, a complexity particularly salient for some types of online research. The ESRC FRE takes no account of these complexities in its discussion of consent and assumes a model of physical co-presence or material exchange:
‘typically the information should be provided in written form, time should be allowed for the participants to consider their choices, and the forms should be signed off by the research participant to indicate consent.’ (2010:pg.28)
4.4 In contrast, the AoiR guidelines reiterate the importance of informed consent but highlight particular issues facing online research that need to be considered, including the medium through which informed consent is negotiated, the role of moderators or list owners as gatekeepers, possible uses of data, confidentiality and anonymity issues and the legal and ethical requirements of disciplines and countries implicated in the research. The AoiR guidelines recognise that informed consent needs to be approached in flexible, emergent and adaptive ways:
‘In some contexts, however, the goals of a research project may shift over time as emerging patterns suggest new questions etc. Determining not only if but when to ask for informed consent is thus somewhat context-dependent and requires particular attention to the “fine-grained” details of the research project not only in its inception but also as it may change over its course.” (AoiR 2002: pg 6)
4.5 This positioning of informed consent as an ongoing negotiation, rather than a signature on a form at the start of the research process, is particularly important in some internet (and other) research contexts. Research in online environments, often with large and fluctuating populations and an unknown mix of active participants and invisible lurkers, faces particular difficulties in managing informed consent without disrupting the natural flow of the interaction or significantly altering the nature of the environment or interactions being studied. Possible solutions reported in the research literature include use of gatekeepers or ethical statements in regular postings and/or signatures. However, it is not always practical or possible to gain informed consent from the participants that you interact with online and in some research contexts some level of non-disclosure is inevitable. The AoiR recommendations point to the use of facilitator/moderator or list owners as possible gatekeepers, but also note the need to evaluate the extent to which these ‘permissions’ are sufficient in groups with high population turnovers. Research continues to highlight the emergent and insecure nature of applying ‘offline’ concepts such as informed consent to online settings. While not suggesting that these kinds of research settings and research questions should have carte blanche ethical freedom, it is certainly not appropriate or progressive that all internet research is labelled as potentially ethically problematic.
4.6 In the section dealing with consent in multi-disciplinary projects, the ESRC FRE acknowledges that an ongoing reflexive stance is required to deal with the demands of disciplinary and methodological diversity, noting that consent demands a degree of fluidity and is not a simple or linear process. The need for ongoing ethical revision is highlighted and there is the suggestion that ‘highly formalised or bureaucratic ways of securing consent should be avoided’ (2010: pg 29). However, this recognition is not extended to internet research and its inherently multi-disciplinary nature is seemingly discounted. In answering the Frequently Asked Question in the FRE ‘s section 2, ‘Why should internet research receive a full ethics review’, the document poses the questions, ‘How easy is it to get informed consent from the participants in the community being researched?’ and ‘What does informed consent entail in that context?’. There is great value in thinking through the implications of these questions and in critically examining the ways in which informed consent, with its foundations in offline research practices, can be translated into different online spaces. The assumption that all internet research will necessarily require a full ethics review, however, over-homogenises the field, does not move ethical debate forward and as Stanley and Wise (2009: pg.9) suggest may result in entirely non-contentious research requiring a full ethical review simply by virtue of employing some form of internet methodology.
Public/Private: Ethical Expectations and Assumptions Online
5.1 Traditional ethical distinctions of public and private are difficult to translate to social research on the internet and the ESRC FRE questions what constitutes privacy in an online environment and how privacy can be ethically defined and managed. The blurred boundaries of public and private spaces and interactions online and, crucially, individuals’ expectations of privacy in different contexts ,are problematic and shifting constructions. The AoiR guidelines recommend that researchers evaluate the ethical expectations/assumptions of the authors/subjects as well as the status of participants as either ‘subjects’ or as authors of texts intended to be public (AoiR 2002:pg 7).5.2 Online forums or technologically-mediated spaces in the public domain (with public meaning a space that does not require membership or access to be granted by a gatekeeper) may feel ‘private’ to participants and they have been shown to foster a sense of intimacy, community and an expectation of privacy among members. However, there is little consensus on contested and highly contextual understandings of what expectations of privacy mean in different online settings. Similarly notions of ‘reasonable expectation’ are complicated by conflicting definitions of which online environments can reasonably be considered to be publicly accessible. The constructed and shifting nature of these understandings pose interesting questions for the social researcher and throw up ethical challenges that lack exact offline analogues. The questions posed by the AoiR recommendations provide researchers with ways of thinking through potential ethical pitfalls, in contrast to the context-blinkered approach of the roughly sketched FRE.
5.3 The FRE provides a tentative definition of data considered to be in the public domain by classifying information posted in online forums or spaces as ‘intentionally public’, with the caveat that “the public nature of any communication or information on the internet should always be critically examined.” (2010: pg.11) This kind of critical examination is ongoing in the research literature and notions of privacy can be understood as evolving and contingent. The ESRC definition of ‘intention’ is at least as problematic as the notion of ‘reasonable expectation’ that internet researchers have been grappling with for some time, and it does little to clarify an appropriate stance on the ethical accessibility of public/private data. Some online environments are public, accessible and permanent, while other more fluid, and temporal spaces can be ambiguous and ephemeral, each demanding different ethical responses and posing different ethical challenges. Participants’ expectations of these settings are also mercurial and there is evidence to suggest that people’s understandings of their online activities and audiences are not in sync with the realities of online networks:
“While some participants have an expectation of privacy, it is extremely misplaced. More fruitful efforts might be made in educating the public about the vulnerability of internet postings to scrutiny – an inherent aspect of many Internet venues – than by debating whether or not such scrutiny should be sanctioned in research.” (Walther 2002: pg 207)
5.4 These are debates that extend across a range of substantive interests and disciplines, but while the FRE acknowledges the growing literature on online research ethics, the ethical landscapes being mapped by this work have been flattened. More usefully, the AoiR ethical guidelines suggest that in spaces where participants assume or believe that their communication is private, then there is a greater obligation on the researcher to ensure privacy through confidentiality, anonymity and the mechanisms of informed consent. Conversely research drawing on public archives, public web pages and posts to public lists or groups may have a reduced expectation of privacy and a reduced ethical obligation to protect it (AoiR 2002: pg 7). In making this distinction, the AoiR guidelines again emphasise different ways of conceptualising online research that go beyond the FRE definition of research ‘through’ the internet. The need to think about online environments and interactions as texts as well as social spaces and of participants as ‘authors’ as well as ‘subjects’ is not a distinction included in the FRE. Public and private spaces online are both channels for the flow of information, but they differ in the ways they are managed and understood by participants. The assumption of the need for ethical protection is arguably more suited to the risk categories identified by the FRE, such as research involving sensitive topics or vulnerable groups, than to online spaces inhabited by reflexive participants who are making conscious decisions about their own public/private visibility and who have agency over their own online presence and authorship.
5.5 It is important to remember here that research is not always incompatible with participants’ expectations and perceptions of a space and in a web 2.0 environment of converging cultural and media forms, and an increasing emphasis on internet users as authors or user-generators, ethical concerns about public/private boundaries have to be evaluated in context. For example, research that uses respondents’ own videos or blogs as a data collection tool may actually be ethically empowering in demarginalising participants’ voices, while online questionnaires may engage respondents who would not be willing or comfortable participating in research on a face to face basis.
The Othering of Internet Research?
6.1 Informed consent and the public/private distinction are often focal points for debates about internet research ethics and are issues that pose particular ethical problems for internet research. The online field often lacks exact analogues with offline research ethics and settings; but this is not an argument for bracketing it off entirely and there is much evidence to suggest that online research is simply facing old methodological concerns in new spaces.6.2 Issues of confidentiality and anonymity, data storage, research risks and ethical differences in global and local contexts, may be ethically problematic elements of internet research. However, this is not so as a given and it may also be true of a number of traditional research settings and methodological approaches. Situating informed consent, deception and covert observation as key ethical concerns also penalises qualitative research and its potential uses of textual and visual data, suggesting that quantitative methodologies are somehow less problematic or immune from online ethical breaches; what Walther (2002) has called a methodological myopia. This matches the myopia of the FRE blanket definition of internet research as by nature risky, which fails to take the critical and adaptive stance advocated by AoiR and which conflates a range of qualitative and quantitative methodologies and approaches to research, analysis and dissemination. In so doing, the FRE is in danger of demonising online research methods and inflating ethical issues in online research. Like offline research settings the internet is a differentiated and heterogeneous social and cultural space and framing ethical conduct in this way may result in Research Ethics Committees asking for things that are irrelevant or impossible for internet research. More important to the development of online methodologies is empirical work that explores the complexities of ethical research in new technocultural environments and identifies types of internet research where careful ethical attention is required, as Thomas argues:
‘ethical conundrums are never easily solved, and dialogue, critique, constant vigilance, and accountability seem far preferable to more rules and increased oversight.' (Thomas 2004: pg 200)
6.3 The FRE states that concerns around internet research in the consultation process resulted in the recommendation for a full ethics review for this type of research, but it also notes the rapidly evolving and developing nature of internet research as an area. Beer and Burrows (2007) argue that issues of trust, privacy and surveillance are central to understanding a web 2.0 society and advocate research which actively engages with the medium of its study:
“In order to get some idea of users and their practices it is necessary to become a ‘wikizen’. The social researcher will need to be immersed, they will need to be participatory, and they will need to ‘get inside’.” (Beer and Burrows 2007: pg 10)
6.4 Rather than contributing to lively debate and tackling some of the emergent, complex and contested understandings of what ethical online research might look like, the ESRC FRE framework unhelpfully proposes a formal review structure that is reminiscent of early moral panics around the potential social impacts of new technologies.
Notes
1 <http://aoir.org/reports/ethics.html>
References
ASSOCIATION OF INTERNET RESEARCHERS (2002) Ethical Decision-Making and Internet Research: Recommendations from the AoIR Working Committee, <http://www.aoir.org/reports/ethics.pdf>BEER, D. and Burrows, R. (2007) ‘Sociology and, of and in Web 2.0: Some Initial Considerations’. Sociological Research Online <http://www.socresonline.org.uk/12/5/17.html>
ESRC (2010) Framework for Research Ethics Swindon: ESRC <http://www.esrcsocietytoday.ac.uk/ESRCInfoCentre/Images/Framework%20for%20Research%20Ethics%202010_tcm6-35811.pdf> (accessed October 2010)
STANLEY, L and Wise, S (2010) ‘The ESRC’s 2010 Framework for Research Ethics: Fit for Research Purpose?’ Sociological Research Online, Volume 15 Issue 4 <http://www.socresonline.org.uk/15/4/12.html>
THOMAS, J. (2004) ‘Re-examining the ethics of internet research: Facing the challenge of overzealous oversight’, in Johns, M. D., Chen, S. S. and Hall, G. J. (Eds.) (2004) Online Social Research: Methods, Issues, and Ethics. New York. Peter Lang. pp. 187-201.
WALTHER, J (2002) ‘Research ethics in internet-enabled research: human subjects issues and methodological myopia’. Ethics and Information Technology. 4, 3, 205-216 [doi:10.1023/A:1021368426115]
