Making sense of ‘slippages’: re-evaluating ethics for digital research with children and young people

ABSTRACT

In this paper, we argue that institutional ethical procedures do not properly prepare children’s geographers to conduct digital research with children and young people (CYP). To address this, we propose a relational, dynamic approach to ethics fit for such contemporary research with CYP. We find fault in a growing disconnect between processual expectations of our academic institution and the realities of ‘ethics-in-practice’. This is especially pronounced in an age where digital technologies have created fresh challenges for researchers. As a result, when ethical ‘slippages’ occur – moments departing from expectations but not leading to clear breaches of ethical conduct – researchers are left ill-equipped to manage situations that arise. We reflect on our own experiences of researching digital technologies in youth work settings in the UK to present two examples of ethical ‘slippages’, before offering our reconstructive propositions for ethical procedures fit for digital children’s geographies research.

KEYWORDS:

• Ethics
• children
• young people
• agency
• youth work
• technology

Moments, dilemmas and ‘slippages’

No matter how well we prepare for entering the field, we can never avoid the complexity of ethical challenges, which reveal the gulf between ‘what is’, ‘what should be’, and ‘what our organisational guidelines instruct us to do’. These tensions gain significance when working with children and young people (CYP) as it is critical to protect participants’ anonymity, confidentiality and safety (Tisdall and Kay 2017). Yet we are also deeply aware of the inadequacies of institutional ethical processes for preparing researchers for the ‘real world’ of social research with CYP (Beazley et al. 2009). We argue that a gap exists between procedural ethics – the bureaucratic classification process of seeking ethical approval for research – and ethics in practice – the ‘day-to-day ethical issues that arise in the doing of research’ (Guillemin and Gillam 2004, 264). This doing of ethically sound research can often take us well beyond the scope of our procedural ethics, leading to ‘moments’ (Guillemin and Gillam 2004; Holt 2004) or ‘dilemmas’ (Morrow 2008) in the field that encourage us to stop and think about the ethical imperatives that are in tension.

We write this article as two UK-based PhD researchers of digital technologies in youth work settings. For us, digital technologies are not just about devices or applications, but also the broader sociotechnical infrastructure that underpins these, such as cheap mobile data, freely available Wi-Fi and the prevalence of mobile devices. We have experienced the tensions of doing ethical research first-hand, supporting CYP alongside professional youth workers. In doing so, we have faced difficulties in navigating our institution’s ill-fitting ethical procedures alongside the realities of conducting participatory research. Our inspiration for writing this came from an unshakeable feeling that institutional ethical procedures – designed to protect participants and prepare researchers for conducting ethical research – did not account for the realities of research with CYP.

Research with CYP is necessarily fraught with complexities in power relations and ethical obligations (Holt 2004; Skelton 2008), especially when exploring the use of digital technologies. As critical activist researchers, we share an explicit commitment to social justice, working against the various forms of oppression that CYP face in society and centring their rights in our work (Jarvis 2017). In contrast, ethics boards often treat all young participants – whether 1 or 17 years old – as ‘incapable of protecting their own interests’, and classify CYP to be homogenously ‘vulnerable’ (Schäfer and Yarwood 2008). Far from protecting their interests, this reinforces CYP’s oppression (Skelton 2008), and creates problems for children’s geographers attempting to challenge this positioning of CYP.

This lack of connection between procedural ethics and ethics in practice has become more pronounced as digital devices and infrastructures pervade CYP’s lives. We research the role these technologies might play in supporting young people’s participation in society. Whilst we are acutely aware of the fresh ethical challenges that digital technologies present for children’s geographies (Hadfield-Hill and Zara 2018), we contend that the tendency of procedural ethics to centre harms and risks fails to consider the myriad of benefits that these technologies can offer CYP (Ergler et al. 2016). We contend that procedural ethics systems have not kept pace with the complexities of digital research with CYP. This leaves researchers unprepared for overcoming the challenges they may face in the field.

We respond to Elsbeth Robson’s (2018) invitation to reflect on doing ethical research with CYP by problematising the ‘moments’ and ‘dilemmas’ we have faced in our own research. We are inspired by Fanon’s (1986) concept of an ethical slippage (un glissement ethique) to think through the shortcomings of our institution’s ethical procedures. For Fanon, an ethical slippage described the ways that the moral values of white France were able to take hold in the consciousness of black Martinicans (Sullivan 2004). He sought to understand how the values and priorities of one world were transplanted to another world in which they did not fit and led to oppression. We propose that the ethical ‘dilemmas’ experienced in the pursuit of digital research with CYP are a kind of ethical slippage, which show how currently-existing systems are ill-suited for the nature of this research. Making sense of these slippages can help us to understand the gap emerging between procedural ethics and ethics in (digital) practice. We then take up the challenge posed by Hall, Sou, and Pottinger (2021, 50) of critiquing the ‘institutional norms of research ethics’. We argue that procedural ethics exists as a classification system (Bowker and Leigh Star 1999) – so when things exist outside of the classifications set by ethical procedures (such as new forms of digital research with CYP), we experience ethical slippages that draw attention to the ill-suitedness of the classifications. Showing how procedural ethics conditions our research as a biopolitical technology (Foucault 2002) focused on the classification of risk, we go on to offer reconstructive propositions to centre alternate values in procedural ethics systems that advocate for a more meaningful, relational ethics in our research with CYP.

Between ‘vulnerable’ and ‘competent social actors’

In contrast to CYP’s historic positioning as powerless and vulnerable, children’s geographers seek ways of supporting CYP as ‘competent social actors’ (Freeman and Tranter 2011, 224). The greater recognition of CYP’s agency and participation in research – for example, through their active involvement as co-researchers (Hampshire et al. 2012) – departs from the notions of CYP as ‘citizens-in-training’ and instead respects their capabilities at the current stages of their lives (Cockburn 2013). Adapting with the changing nature of childhood, adolescence and youth, children’s geographies have continually questioned what their work is ‘good for’ and how it benefits CYP (King and Priestley 2008; Robson et al. 2009; Freeman 2020). Methodologies have also changed, moving from ‘being with’ CYP and observing their behaviours (Ward 1978; Hart 1979) to ‘doing with’, collaborating directly with groups of CYP through participatory and emancipatory methods to advance social justice causes in the community (Aitken 2018). Researchers now commonly reinforce CYP’s rights to be fully included in society (Beazley et al. 2009), promote reflexivity and dialogue with CYP (Christensen 2004) and support their participation in civic projects (Peacock, Anderson, and Crivellaro 2018). Visible action initiated by CYP has also contributed to this, such as the youth-led school strikes for climate (Jenkins et al. 2016; Collins 2020).

For Beazley et al. (2009), CYP have the ‘right to be properly researched’, meaning that research questions, methods, and ethical procedures should respect CYP’s autonomy, capabilities and diversity of lived experience (Abebe and Bessell 2014). Hall, Sou, and Pottinger (2021) summarise the five key premises of such ethical procedures as autonomy (choosing to take part), beneficence (ensuring the research has social benefit), avoidance of harm, confidentiality and anonymity, and integrity (managing conflicts of interest). However, Morrow and Richards (1996) distinguish research with CYP by the need to respect their differing competencies and be mindful of the disparities in power and status compared with adults. They also warn against viewing CYP as a homogeneously vulnerable and incompetent group and draw attention to their right to a voice in research, enshrined in the UN Convention of the Rights of the Child (United Nations 1989). Respecting the rights of the child and adopting a CYP-centred perspective is a model approach we should aim for (Bell 2008; Bell and Members of the Glasgow Centre for the Child and Society 2008).

Despite this, architects of ethical procedures have codified assumptions about CYP and their inability to participate in research on a level with adults. The requirement to obtain ethical approval from an institution to conduct research reduces the scope to challenge such assumptions in the field. We recognise that valuable work has gone into making these procedures more relevant for contemporary research (Alderson 2012). Yet the historic emphasis has been on creating ‘a multi-layered bureaucratic process … rather than promoting a set of supported professional values and behaviours’(Wilkinson and Wilkinson 2019, 481). Several scholars have proposed redesigning ethical procedures to account for the realities of working with CYP (Pain 2008; Skelton 2008). Such relational and dynamic approaches to informed consent, participant confidentiality and risk would better support participatory researchers to design research that reflects the changing nature of CYP themselves (Beazley et al. 2009). However, these proposals represent a significant challenge to the biomedical foundations and institutional values driving the production of ethical procedures. These include emphases on reputational management, the primacy of written parental consent and CYP’s vulnerability to harm over the benefits of their participation (Skelton 2008; te Riele and Brooks 2013).

New pressures and opportunities of digital technologies

The contemporary youth experience in the UK is now shaped by the pervasive availability, adoption and use of digital technologies (Plowman 2016). This cultural shift provokes ethical questions that have not yet been accounted for in procedural ethics systems, and for which there are often ill-defined standards. For example, a traditional approach to procedural ethics would mandate that if we were to use social media posts from ‘participants’ as data, we should take every effort to anonymise it (Dawson 2014). However, the data might well already be public, rendering any efforts to anonymise ineffective. von Benzon (2019) argues that such amateur online authors are informed agents, who often put their writing online publicly in the hopes that they are read. Here, anonymisation could thus strip them of their agency and intent. This issue and others like it have not yet been accounted for within procedural ethics systems, which remain grounded in pre-digital understandings of ethics.

With digital technologies and human geographies now ‘intricately intertwined’ (Duggan 2021, 253), children’s geographers have begun engaging with the ethical questions arising from digital technologies in earnest. The blurring of the lines between online/offline and public/private has created fresh difficulties for researchers, as they try to negotiate new understandings of privacy and consent while maintaining an ethical research practice (Ergler et al. 2016; Palfrey and Gasser 2016). Many CYP, for example, openly share personal data on social media – a practice that invites many ethical questions. Conducting research with and through technology also invites us to reconsider the tools that we use (Hadfield-Hill and Zara 2018). For example, Volpe (2019) has described the myriad of privacy issues inherent in using PhotoVoice methods in a digital context, which can include the unintentional inclusion of individuals who have not consented to participate in the research project. Similarly, whilst social media and blogging can help CYP with identity formation, these tools come with many risks, such as exposure to harmful misinformation (Hawkins and Watson 2017).

Digital technologies can offer novel opportunities to engage in life-affirming practices to ease the passage through adolescence. Wilson (2016) describes multiple instances of care-experienced CYP engaging in ‘flexible self-identification’ that is mediated through digital technologies as a tool of self-care. In this research, some CYP used video games as a way of processing stress, while others negotiated difficult situations with music played through their iPhones. Digital technologies also offer novel methodological opportunities, such as the ease of conducting participatory video research, which gives CYP agency over how they represent their lives (Wilkinson et al. 2020). Procedural ethics systems do not yet reflect these experiences, tending towards an understanding of technologies as inert artefacts (e.g. Jones, Williams, and Fleuriot 2003) rather than a mediator or assembler of experiences. As such, our current procedural ethics systems are ill-suited to advise on the ethics of technological work. We argue that this gives rise to ethical slippages that leave researchers inadequately prepared for the realities of digital research with CYP.

Ethical encounters

Two examples of a ‘slippage’As PhD students, we acknowledge that there is a degree of training and skills development inherent in our work (Mowbray and Halse 2010; Pitt and Mewburn 2016). Although research training may touch upon ethical issues, we feel that it rarely touches upon the realities of ethics in practice or the specific challenges of doing research during a time of digital transformation. By only focusing on limited aspects of the ethics of research, we believe that institutional ethics systems invariably contribute to the proliferation of slippages – between ‘what is’, ‘what should be’ and ‘what our organisational guidelines instruct us to do’. We hope that by presenting two slippages we have encountered, we can connect these theoretical points to lived experience, and help others to resonate with the dilemmas we faced in our research (Morrow 2008). To protect the privacy and confidentiality of those taking part in our research, we use pseudonyms in place of the real names of organisations and participants throughout our slippages. First, we discuss our positionalities and our relationship to the work we present (Banks and Brydon-Miller 2018).

Kieran

I’m a sociologist and designer who does research with charities and care-experienced CYP. I was initially drawn to my work to personally ‘make a difference’, but I have since realised that often, the best work I can do is to communicate, facilitate and translate what is already happening. I am a white man, and I try to remain aware of how my whiteness and masculinity affects the work that I do. My research currently focuses on charities after austerity, and developing methods to support sustainable collectives of critical young people and frontline charity workers. Researching in this context, I couldn’t help noticing ethical slippages everywhere I went.

Sean

I research how CYP can be more involved in shaping the places around them. From my background as an urban planner, I became aware of how those responsible for shaping our built environment do so without involving marginalised people. CYP are often an afterthought for planners, and people are quick to make assumptions about their motivations and abilities. This needs to change. I recognise I hold a position of privilege, as a white, cis, adult male conducting research in a UK university. Driving my work is a desire to use the privilege I hold to improve the situation for those currently unheard, and wielding the democratic power of technologies to achieve this.

‘Does she know that we now know?’: digital disclosure

Kieran

I conduct research with Building Bridges, a project that aims to connect care-experienced young people and their youth workers across the UK. In November 2018, I became part of the project’s facilitation team, helping to plan residential weekends while conducting research with workers and young people. I was part of two WhatsApp groups created by the project team to keep in contact between residentials – one for the project’s workers, and one for the young people. I had no specific concerns about this, as the workers and I had experience of managing young person-centred WhatsApp groups before. As groups were travelling to the second residential, Amy, a worker, sent a message to the workers’ group chat:

[10:12]

Amy: Scarlett has disclosed she's pregnant just before we got on the train

[10:14]

Frank: How’s she feeling about it?

[10:14]

Karen: Thanks for letting us know. Can we have a chat when you arrive? Thanks x

[10:15]

Jess: Does she know that we now know? Should we wait for her to tell us?

[10:16]

Amy: She's not happy about it. She doesn't feel it's the right time and she used contraception. Best not to say anything to her for now – she’ll say something if she wants, but think she wants to do some processing this weekend. I just thought I'd tell you for health and safety reasons during the residential.

Building Bridges has various safeguarding, health and safety procedures that require knowing a lot of personal information about participants, including detailed knowledge of their physical and mental health. Being privy to this highly personal disclosure hours before a residential begins throws all semblance of established procedure into disarray. Immediate thoughts turned towards practical realities – how can Scarlett be kept safe? What activities can Scarlett do? How do the workers and I deal with the fact that everyone now knows about Scarlett’s pregnancy when Amy has asked us not to say anything? Making the wrong decision could easily break a young person’s trust. This is the sort of ethical decision-making that ethical processes do not prepare you for.

Amy had followed the correct disclosure and safeguarding process, but by doing so on the workers’ WhatsApp group, she had made everyone responsible for how we acted on that information. What happened next posed an even greater ethical dilemma. Aided by the instantaneous nature of WhatsApp, and not knowing that Scarlett did not want to share this news with the group, other workers who were travelling with young people to the residential shared the news with young people they were with. However, this was before Amy’s second message arrived four minutes later, where she had asked everyone not to share it. At this point, nobody had any way of knowing who knew about Scarlett’s pregnancy, and who knew that it should remain confidential. By trying to ensure our groups had a strong digital infrastructure to communicate and build bonds between residentials, the workers and I had facilitated a slippage – we had done what was ‘right’ according to procedure, but created a situation that was out of our control and could lead to a young person feeling betrayed.

All workers were well trained in methods of de-escalation, yet many of us still had to spend the weekend pretending not to know about Scarlett’s pregnancy. This became personally problematic at one point, when Scarlett was experiencing morning sickness and she asked if I could buy her a bottle of water. I felt uneasy about treating a participant differently to others, but Scarlett had assumed that I knew about her pregnancy and wanted my help. I could have followed procedure and told her I would not buy her the water, or that she had to pay, but I knew that would not be ‘right’. While I bought her the water and sat with her for a while, another young person asked where Scarlett got the drink from and whether we could get them one. Luckily, Scarlett explained the situation to the other young person, but every moment of that conversation was so ethically charged that I could not be sure I was doing something ‘right’. These are small and even mundane decisions, but when trusting relationships count for so much, they matter.

The digital disclosure that Amy shared and the subsequent actions of youth workers had dropped me into a set of slippages where it was much easier to make a mistake than ‘get things right’. By doing so, I was forced to exist outside ethical procedure, but within my own ethical practice and feelings of justice. Procedural ethics might speak of the content or process of a disclosure like this, but says nothing of the actual ethical responsibilities we owe to the people we work with – particularly when they may have fewer material resources and greater needs for relationships built on trust and care. The lead facilitator reassured me: ‘Safeguarding be damned, you’d be a bad person if you didn’t buy her that drink’.

‘Pics or it didn’t happen’: digital anonymity

Sean

I take the duty of care that I have to my participants incredibly seriously, going to considerable lengths to protect CYP’s anonymity (Bell 2008). The collaborative nature of my research, however, continually draws this duty into question. From April 2018 to December 2019, I immersed myself in the machinations of Bramwell Youth Council, a youth group composed of 11-18-year-olds and supported by several youth workers. I spent nine months observing meetings before initiating an embedded action research project. My ethical commitments meant I refrained from recording the whole story of my research setting, leaving out key details that could identify participants. I filled my field diary with pseudonyms and vague descriptions of events I had witnessed. I blurred faces from photos and removed street names, names of schools and other identifiable information from artefacts. The meticulous process of extraction I went through to comply with institutional ethical requirements inevitably led to me stripping some of the richness from my work, particularly as it concerned specific local geographies. Yet this was a compromise I made to protect CYP.

During my time with Bramwell Youth Council, social media began to take on an increasingly prominent role. When I first started working with the group, youth workers used a private Facebook group to check who could attend meetings and events, share photos and keep in touch. One day, the youth workers set up a Twitter account for the youth council and started to post photographs of meetings. Photos of the group, sometimes accompanied by their names, were now entering the public domain with no restrictions on how these were viewed and shared. Details of my research began to accompany these posts as my research progressed. Encouraged by the reception they were getting on Twitter, two youth council members set up their own Twitter accounts, which they ran themselves from their own smartphones. These accounts also began including photos and details of the research, which were then shared widely among the members’ Twitter followers (as is the way Twitter works). Suddenly, I was in a situation for which I was unprepared and could not control. However, I did not want to discourage the youth council members, who were so proud of their work, from doing this. It became clear that I was experiencing a slippage: the preparations I had made through my institutional ethics were insufficient to deal with this wilful breach of anonymity.

This reached a crescendo when local media took an interest in the youth council’s work. As the project developed, the group wanted to make waves in their local community. The project focused on cleaning up their local parks and beaches – so persuading residents to join in their efforts and look after their city was key to the project’s success. I observed the youth workers and youth council members excitedly reaching out to local press outlets to publicise their efforts, and a local newspaper obliged. Instead of writing a story about the whole group, the newspaper singled out a member of the youth council for their creation of a striking poster to display in shop windows. This article was published both online and in print, with a large headshot photo of the young person accompanying details of where they lived and went to school.

This personal information was now in the public domain. My ethical approvals were redundant, but I was too close to the end to modify them. Perhaps the more significant ethical conundrum was that while this felt procedurally incorrect, the chain of events felt ethically correct. If this is what participants and youth workers wanted, who was I to call it unethical? Participants not only consented to this, but also actively sought to share their efforts with as many people as possible through the various Twitter accounts and local media. All of this was encouraged by the youth workers, who prepared official press releases to be issued by the organisation and tweeted these across their professional network. I even relented myself when I retweeted it from my personal Twitter account.

My pride in Bramwell Youth Council’s work was tempered by an uneasiness in undermining my institution’s ethical procedures. In this slippage, my institution’s ethical procedures repeatedly fell short of the realities of activist research with CYP. To me, the ethical thing to do was to respect the wishes of the participants I worked with, and spread the message of their activism. Conversely, I felt it would be unethical to try to stop or slow them and would be completely at odds with the rights-based agenda that underpinned my research. How can I claim to support CYP in getting their voices heard if I was the one muffling them? I could not have accounted for this potential in my ethics application, as procedural ethics would not know how to deal with such a grey area. Ultimately, this slippage meant that it was extremely difficult – perhaps even undesirable – to follow ethical procedure.

Kieran and Sean – joint reflections

The slippages that we have described find a common source in the static and rigid expectations of procedural ethics systems: procedural ethics set expectations, and we fail to meet them because of the realities of our research practice. In the first case, digital forms of disclosure led to a chain of events that did not have clear precedent, leaving us in doubt about the ‘right’ course of action to take. In the second case, the research gained a life of its own on social media in a way that went beyond the scope of what an ethics form could capture. However, both involved compromises to the anonymity of our participants and were counter to our clear-cut ethical procedures – but the extent to which we could have (or should have) prevented these situations from happening is unclear.

Despite the support of supervisors, colleagues, and peers in helping us to understand the requirements of institutional ethics, slippages are personal, subjective and unexpected. Supervisors can play an important role in guiding students to make ethically sound decisions, but in the field, it is our responsibility to conduct our research ethically and handle slippages as they occur. To understand the conditions that facilitate slippages, we felt it important to understand the shortcomings of the ethical review process that had failed to prepare us for these.

Making sense of slippages: reconstructing the ethics form

Although institutions have diverse approaches to ethical procedures, the entry point for most researchers is some kind of ethics form. Early into this work, we were struck by the strangeness of how our colleagues spoke about these forms: ‘I’ve done my ethics’, or ‘I’ve got my ethics’. While our colleagues were referring to the transactional experience of obtaining ethical approval through the ethics form, we were interested in interrogating the work of the ethics form.

We now critically analyse of our institution’s ethics form to illustrate its function and disconnection from digital research. We present this as an example, owing to our familiarity with it; however, we are acutely aware that the issues we discuss here are by no means isolated to our institution. Our interest is not to level criticism, but to read between the lines of the processes that make such forms commonplace at academic institutions. What ontological assumptions does it make? What worlds does it try to assemble, and how do these differ from the worlds we try to assemble in our research? And how does it facilitate ethical slippages? Through our analysis, we began to see the ethical review process as a technology of biopolitics – the processes of power which aim to ‘regulate and control life’ through the classification of ‘legitimate’ knowledge (Foucault 2002; Lemke 2011).

The ethical approval process at our institution and many others hinges upon two central components: completion of an ethics form, which flags proposals as either ‘low risk’ or ‘high risk’, and review of ‘high risk’ project proposals by an ethics committee to establish whether they achieve ‘ethical approval’. Both stages are binary classification systems: projects are either ‘low risk’ or ‘high risk’, ‘ethically approved’ or ‘not ethically approved’. The process does not seek to establish how ethical a project is, just its risk and approval status. There is no facility here, for example, to track projects that may encounter more issues and keep in contact with these researchers. Instead, by operating with these binary classifications, our ethical review processes seek to limit the range of possible action taken by researchers.

Bowker and Leigh Star (1999) posit that classification systems (such as the ethics form at our institution) are contingently created and have material impacts. In this case, whether a project is ‘low risk’ or ‘high risk’ affects its approval status, which subsequently affects the possibility of undertaking research. ‘Low risk’ status leads automatically to ‘ethically approved’ status (a computer-generated confirmation email). ‘High risk’ status leads to a halt being placed on the project and the start of a process of manual scrutiny by university colleagues to obtain ‘ethically approved’ status. We are familiar with colleagues who have waited six months for this approval and, as stipulated by the form, have been left unable to do any research during this time. By pausing the research process like this, the ethical review process acts biopolitically and determines what it considers ‘legitimate’ and what it considers ‘illegitimate’. Here, the tension is that ‘high risk’ research is illegitimate until it has been deemed sufficiently ‘low risk’. As researchers who work closely with CYP and others who are considered vulnerable or oppressed, we fall into this category automatically. By proposing to work with digital technology in this same context, we only serve to intensify the risk in the view of the ethics form.

At our institution, research with anyone under 18 years old – regardless of the questions or methods – is classified as ‘high risk’. The explanation given for this is because the form conflates children with other vulnerable groups and believes these to be ‘incapable of protecting their own interests’. Ensuring the safety of CYP in our research should of course be a priority. Yet the form’s role as classification technology means that it pays little attention to the work done to ascertain CYP’s competence. For example, it contains no mention of assessment of Gillick competency, which recognises the ability of practitioners to assess whether a child is able to consent for themselves, depending on their maturity and the nature of consent involved (NSPCC 2020). Instead, it erases the possibility of agency for CYP, seeing them only as risk factors. It does not account for diverse experiences, nor the ways that different CYP will use and appropriate digital technologies: it seeks only to classify risk.

The form also asks how many participants we will recruit (with the form only allowing an exact number), whether they will be subject to ‘prolonged or repetitive testing’ and refers to the principal researcher as ‘lead scientist’. This makes it clear the form is more appropriate for those conducting positivist forms of research. We often do not know how many people will be taking part until we are at a session, as we conduct research with organisations with drop-in attendance policies. While we do conduct research with CYP for hours at a time (e.g. at weekend residentials) and have spent months building trust and rapport with participants, this is not the same as ‘prolonged or repetitive testing’ as medical research would imagine it. Defined as more than four hours commitment or attendance on more than two occasions, this would leave us falling foul of this criterion, despite the terminology and implication bearing little resemblance to our work. Crucially, the ethics form works on the positivist assumption that it is possible for researchers to have a priori knowledge of all possible details of the research to be undertaken – particularly all risk factors. This element of unpredictability can increase if conducting research online, as it becomes even more impossible to be able to enumerate all risk factors because of the scale and complexity of the internet. Whilst we may seek to protect our participants, we cannot entirely shield them from the forces of the internet in all situations.

Other risk factors stipulated by the form include discussion of sensitive topics, or the possibility of psychological stress or harm. We often work with CYP with traumatic experiences. For us, it is frequently impossible to know what constitutes a ‘sensitive topic’ or what might cause psychological harm for the specific CYP that we work with until we meaningfully know them. Yet the approval process dictates that ‘ethical approval’ must be granted before any fieldwork takes place. This precludes participatory or action-centred research approaches, which necessitates cycles of dynamic adjustment and reconfiguration of the research by participants (Stringer 2014). In positivist research, where the object of analysis is a specific, targeted form of intervention, it may be possible to ascertain variables, isolate conditions and enumerate all risk factors; but we believe that such an approach is not achievable or desirable in digital research with CYP. As such, it becomes clear that the binary classification systems of our institution’s ethical approval processes are ill-fitted for forms of research centred on anything other than ascertaining risk. If we wish to conduct forms of research that are considered ‘risky’ (for example, working with CYP or using digital technologies), then we do so against the grain of the ethical approval process, and we are made to experience slippages.

Centring alternate values: reconstructing ethical research

Understanding the ethical approval form as a biopolitical technology that seeks to classify research based on its perceived risk calls attention to the power inherent within the mundane act of filling in the form. By submitting to our institution’s interpretation of ethics, we transform ethical processes that should be focused on morality, equity, justice and liberation, into bureaucratic ones focused on traceable decision-making, classification of risk, and legal compliance. If the ethical approval process bears such little relation to the realities of situational ethical decision-making – ethics in practice – then we must begin to conceptualise some form of egress from it (Fisher 2016). We need a system that holds greater relevance for the creative, engaging and life-enhancing (Stringer 2014) research that occurs in children’s geographies and beyond, and which account for the slippages that we have experienced. Other scholars have begun this valuable work by proposing forms of participatory ethics. These propositions are grounded in the idea of designing ethical procedures that do not contribute to CYP’s further oppression. This may involve co-creating ethics iteratively with CYP, ‘responding to their feelings and beliefs about what is ethical and how they would like to be treated’ (Pain 2008, 107) as well as with researchers at all levels whom enact these procedures in the field (Skelton 2008). We build on these constructive attempts by proposing a centring of values other than the risk factors that are so prominent in our current ethical approval processes. Here, we choose to centre the values of care, justice, and agency, which we believe are not only vitally important, but more amenable to the conduct of digital technology research.

Analysis of our institution’s ethical approval process showed an assumption that researchers are conducting positivist research that attempts to maintain the facade of distanced, impartial observation. Yet as Barad (2007, 47) has shown, there is a ‘taken-for-granted ontological gap’ at the heart of the naive realism of positivist approaches. Knowledge is a representation of that which it claims to know – but there is always a knower on the other side of this process producing knowledge in the first place. Put simply, we cannot conduct impartial observation because we produce that which we observe. To refocus this in a generative sense, we follow de la Bellacasa (2011, 90) in arguing that laying these relations out clearly can serve ‘a gathering purpose’ capable of holding together assemblages of relations. If we centre matters of care in our procedural ethics systems, then we might ask questions about researchers as producers of knowledge, focusing on their positionalities and experiences. These could include questions such as: what is your experience of conducting research that features these ethical issues? What is your interest and intention with this work? What ‘gathering purpose’ are you trying to serve?

For de la Bellacasa (2011, 99), bringing matters of care into procedural ethics systems ‘is a way of relating to them, of inevitably becoming affected by them, and of modifying their potential to affect others’. The a priori perspective that the ethical approval process encourages us to take also promotes methodological approaches that do not change throughout the life of the research. Yet methodologies that focus on participatory research, emergent exploration and iterative cycles of action such as grounded theory and participatory action research have had currency for decades. Designing ethical systems that better account for participatory forms of research would be a start, but we are simultaneously aware of how such participatory methods may be co-opted – particularly within research with CYP (Cooke and Kothari 2001). Rather, we suggest going further to foreground justice as a value within procedural ethics. An orientation towards research justice necessitates research that works to transform structural inequities and facilitate lasting change (Jolivétte 2015). Integrating research justice into ethics systems asks us to constantly question for whose benefit we conduct our research, and how we can promote healing and repair from past and present structural inequities through our work.

If current procedural ethics focus almost exclusively on the classification of risk, then the most subversive value we can centre as an alternative form of research accountability is distributed agency. Drawing on the work of Bennett (2010), we could move from the use of ethics systems to assign ‘singular blame’ to recognise the co-construction of research outcomes, and instead envision alternative ethics systems that acknowledge the agency of all actants within a particular research phenomenon. This might include thinking about the agency of CYP, gatekeepers and us, along with the technologies they use, the socio-political practices oppressing CYP and the nonhuman aspects of their worlds, to name a few. We could think about how harms and benefits are distributed through the conduct of our research, how these intersect with the positionalities of our participants and their experiences of oppression, and how we can ensure the maximisation of collective agency throughout.

Towards a more relational ethics for digital research with CYP

In this paper, we have reflected on the ethical challenges we have faced as digital technology researchers through the lens of slippages and tried to understand their source through a deconstruction of our institution’s ethical processes. The prevalence of digital technologies has posed new ethical questions to social researchers and demands changing systems. As we have shown, the tightly bounded systems of procedural ethics do not accept any possibilities of difference, meaning that further cultural shifts will add further pressures. If a classification system is constructed in one context and remains inflexible, it makes the classification system even less relevant than before (Bowker and Leigh Star 1999). For us, this leaves procedural ethics even more distant from ethics in practice (Guillemin and Gillam 2004).

In centring alternate values, we have shown the possibility of alternate ethical systems. We have only sketched an outline here, but tightly bounded or binary classification systems that seek to position research as ‘ethical’ or ‘unethical’ are insufficient for the conduct of digital technology research with CYP because it always deems CYP and digital technology a ‘risky’ combination. Our experiences of slippages suggest to us that it might be generative to centre values such as care, justice or agency. However, it may be possible to centre any number of alternate values – either to construct a more appropriate procedural ethics, or simply to imagine what a system centred on these values might look like. We stress, however, that we are not calling for an end to safeguarding practices that seek to protect CYP from harms. Rather, we suggest that it is still possible to safeguard appropriately while focusing on values other than risk.

Appreciating the biopolitical elements of procedural ethics, our reflections on these ethical slippages have shown that we need an approach to ethics grounded in an understanding of complex and constantly changing field sites and participants. We encourage researchers reading this to consider different imperatives – for example, how the context of UK austerity has facilitated slippages, something close to our research but out of scope in this paper. We also expect slippages to occur following the significant sociocultural changes brought about by pandemics, such as COVID-19. With physical distancing, digitised social and working lives and grave socio-economic prospects al prevalent at the time of writing – and the prime moral imperative being to keep ourselves, and each other, safe – what slippages do we not yet understand that may demand changes to our ethical procedures?

We call not just for a change in procedural ethics, but also for researchers to engage in the rich description of ethical slippages in their research outputs. We cannot identify the inadequacies of our current systems if we do not speak openly about them.

Acknowledgements

We wish to thank the young people, youth workers and organisations for their participation and support in carrying out this research. This research was funded through the EPSRC Centre for Doctoral Training in Digital Civics (EP/L016176/1). Data supporting this publication is not openly available due to ethical considerations. Access may be possible under appropriate agreement. Please contact the authors for more information.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research was funded through the EPSRC Centre for Doctoral Training in Digital Civics (EP/L016176/1).