Ethics in Context: Essential Flexibility in an International Photo-Elicitation Project with Children and Young People

Existing literatures have discussed both ethical issues in visual research with young people, and the problems associated with applying 'universal' ethical guidelines across varied cultural contexts. There has been little consideration, however, of specific issues raised in projects where visual research is being conducted with young people simultaneously in multiple national contexts. This paper contributes to knowledge in this area. We reflect on our experiences of planning and conducting the International CYCLES project involving photo elicitation with young people in Bangladesh, Brazil, India, Japan, New Zealand, South Africa and the UK. While some issues such as varying access to technology for taking and sharing photos and diverse cultural sensitivities around the use of photography were anticipated in advance, others were more unexpected. Balancing the need for methods to be appropriate, ethical and feasible within each setting with the desire for sufficient consistency across the project is challenging. We argue that an 'ethics in context' approach and an attitude of 'methodological immaturity' is critical in international visual research projects with young people.

Keywords: Photo-elicitation; children and young people; ethics; methods; international

CYCLES (Children and Youth in Cities – Lifestyle Evaluations and Sustainability) is a study of young lives in seven cities. CYCLES explores what conditions enable and constrain young people in living environmentally sustainable and fulfilling lives. The project involves local researchers in seven international cities. The diversity of the cities allows us to research the challenges and opportunities for young people's sustainable consumption across various dimensions of difference. Our project includes the mega cities of Dhaka (Bangladesh), São Paulo (Brazil) and New Delhi (India) where urban infrastructure struggles to support growing populations; Yokohama (Japan) a planned 'eco-city'; Christchurch (New Zealand), a city in transformation after devastating earthquakes; Makhanda (South Africa) a small city facing high rates of youth unemployment and deprivation exacerbated by the legacy of apartheid; and Lambeth (UK) a diverse, densely populated London borough.

At the time of writing, the project is still in progress and has three research phases: a desk based review; photo elicitation and a survey. In this paper we focus on the photo elicitation phase, reflecting on our experience of conducting this kind of work with young people in an international context. While ethical and methodological issues around the use of photography in research with young people are well rehearsed, particular issues arise in photo elicitation projects with young people across diverse national contexts.

CYCLES focuses specifically on 12–24 year olds living within cities. Nearly half of the world's population are under 25, and 7 in 10 young people are expected to live in urban communities by 2050 (United Nations Children's Fund, [41]). Moreover, while cities occupy only 2 percent of the Earth's terrestrial surface, they use over 75 percent of natural resources (United Nations-Habitat, [43]) and are major contributors to greenhouse gas emissions (United Nations Environment, [42]). Creating a more sustainable urban future is highly challenging. While cities can be vibrant centres of new ideas, employment, entertainment, and cultural diversity they are also increasingly socially unequal, with many young residents disproportionately exposed to risk, unemployment and material deprivation (Herzer, [19]; United Nations Children's Fund (UNICEF), [41]). Young people are likely to view and experience the spaces and services of the city differently from adults (Matthews & Limb, [31]; Rasmussen, [36]) and engaging with their perspectives is critical.

Sustainability and sustainable consumption are contested concepts (Jackson, [23]). The broad understanding we work with combines concern to minimise environmental harm with a commitment to fostering social wellbeing. The ultimate challenge of sustainable consumption consists in understanding how it might be possible to live well – within the limits of the planet (Jackson, [24]).

While there is extensive literature on young people's consumer culture, to date the sustainability implications of their everyday consumption practices remain under researched (Collins & Hitchings, [8]). Such research might focus on young people's particular exposure to environmental hazards, their lack of access to environmental goods (Holloway & Valentine, [21]) or the resource implications of their activities and purchases. The initial desktop review of our project outlined environmental conditions for young people in each of the cities. In this qualitative phase we explored how young people spend their time and what they value; gaining insight into their transport, food and leisure practices and the resources these rely upon.

Since the late 1990s, research in the new social studies of childhood has emphasised the need to focus on children and young people as competent beings in their own right (Qvortrup, [35]). In CYCLES, rather than conceptualising children and young people primarily as a vulnerable group, or as potential adults, we consider them as young citizens with significant and distinctive experiences and perspectives on everyday life and consumption. Gallacher and Gallagher navigate a useful position between the conceptions of young people as vulnerable on one hand or fully agentic on the other. They argue for 'a critical rehabilitation of the concept of "immaturity in childhood research"' ([12], p. 511) an approach which treats young people, adults and researchers alike as 'always-unfinished subjects-in-the making' (op cit), always learning and changing, fallible yet full of potential. This conceptualisation informs our orientation to our research particpants, but also to our own role as researchers and the necessarily messy and often uncertain process of research.

Our qualitative work in each city aimed to understand children and young people's everyday lives and consumption in seven distinct cities paying attention to the domains of: food; mobility; home life; leisure/communication; shopping and work and/or study. We also explored what they value most and what constrains and enables their capabilities to achieve that in sustainable ways. We used photo-elicitation to enable grounded discussion of these issues from which broader reflections on opportunities for sustainable consumption and the potential to live well could emerge.

In the following section we discuss some of the particular challenges for research design and ethics in an international project like CYCLES, before outlining the specifics of our approach.

A major motivation for conducting international research with young people is the desire to explore aspects of consistency and contrast between their values, experiences and expectations in different national and cultural contexts. However, as those engaged in comparative research have long recognised (Hantrais, [16]), research methods appropriate in one part of the world may be inappropriate elsewhere, posing challenges for the design of international comparative projects. Hantrais concludes that 'cross national comparative work ... demands greater compromises in methods than a single country focus'([16], p. 4). Rather than seeing such 'compromise' as problematic, as our research evolved we came to recognise it as something to be acknowledged and accepted (Gallacher & Gallagher, [12]), part of the rich process of the international co-production of research.

Gobo ([14]) points to the roots of social research methodology in European and American culture and questions whether research methods which 'have been tacitly proposed as universal' (p. 418) are actually suitable for all cultures. He advocates making research methods 'culturally flexible' and 'glocalising' methodology to suit the local culture' (p. 428). Riessman makes a parallel argument about ethical processes, highlighting the problem of:

'applying 'universal' moral principles that have been constructed (that is derived) in one cultural context, and exporting them without modification, to another' ([37], p. 478).

She calls for 'an ethics-in context approach ... realised in the give and take of research relationships on the ground' ([37], p. 473). While the need to balance flexibility with consistency and coherence always characterises qualitative research (Holloway & Todres, [20]), the challenge is much greater in international projects where fieldwork is being conducted in diverse contexts (Hantrais, [16]).

Ethical issues in research involving photography have received much attention, particularly in relation to projects involving children and young people (e.g. Luttrell, [27]; Wiles, Coffey, Robinson, & Heath, [44]; Wiles et al., [45]). In attempts to ensure anonymity, ethical guidelines usually advocate ensuring that no identifiable young people are shown in photographs (Graham, Powell, Taylor, Anderson, & Fitzgerald, [15]), however researchers often note young people's own desire to have their face shown in photographs and to be credited as authors of their own images (Barker & Smith, [2]; Wiles et al., [44]). Langmann and Pick ([25]) reflect on their experience of conducting research photography in India, suggesting that while the requirement for anonymity is intended to protect individuals, in some circumstances it can compromise their 'dignity'. They draw attention to the way in which some codes of ethics developed in the global north privilege individualist perspectives and suggest that this 'is not necessarily entirely appropriate for research in societies where collectives take precedence over the individual' (p. 713). They suggest that research using photography in international contexts should be guided by Clark et al.'s model of a research ethic based on 'care, compassion and a desire to act in a way that benefits the individual or group who are the focus of of the research' (Clark, Prosser, & Wiles, [6], p. 82).

Flexibility of methods and ethical processes is also necessary in research which involves young people of diverse ages. In CYCLES we adopted the age range of 12–24 years in order to include the experiences of young people through adolescence into early adulthood with recognition of the extent to which age may be understood and experienced variably in different cultural contexts. While the approach of methodological immaturity treats young people, adults and researchers alike as 'always-unfinished subjects-in-the making'(Gallacher & Gallagher, [12], p. 511) this does not imply that all have the same capabilities or freedoms, or that the same research methods or ethical processes will be appropriate for participants of the same age in different cultural contexts.

The logistics of managing CYCLES are complex as lead researchers are based in NZ and the UK, and most of the international partners do not have guaranteed funding for their involvement. CYCLES is funded by the ESRC (the Economics and Social Research Council, a UK government funded public body which provides funding for research in the social sciences) as part of the Centre for Understanding Sustainable Prosperity (CUSP), which is based at the University of Surrey in the UK. The project lead for CYCLES (Bronwyn Hayward) is based at the University of Canterbury in Christchurch, New Zealand. Development of project methodology was shared between researchers at the University of Surrey and the University of Canterbury with input from international partners.

While the ESRC funded fieldwork within the UK and NZ and some international collaboration, no funding was provided for researchers or fieldwork in other countries. In Japan and South Africa our partners (Midori Aoyagi, Aya Yoshida and Ingrid Schudel) are social scientists based within universities or research institutes, while in Brazil and India our partners are environmental campaigners and activists who lead NGOs (Vimlendu Jha leads Swechha in New Delhi and Helio Mattar leads Instituto Akatu in São Paulo)[1]. A partnership with Dhaka was established through funding from the University of Canterbury for a Bangladeshi PhD student (Md. Mehedi Hasan). We were able to provide limited funds for partners' fieldwork expenses and participation in project meetings, but in developing the research design we had to be sensitive to the fact that they were essentially participating in a voluntary capacity. This made us particularly committed to enable them to shape the research in their own city – not only to make it culturally appropriate but also to ensure that it was practically manageable and also interesting and valuable for them given their own agendas and priorities.

Thus, for all of these reasons, it was recognised from the start that we needed to outline shared research protocols and ethical guidelines which could be adapted as appropriate by researchers working within diverse international and organisational contexts.

Asking young people to take photographs which are then used to stimulate conversations in interviews has become an increasingly popular way of eliciting narratives, enabling participants to draw attention to everyday experiences, practices and products of significance to them.

Of particular relevance for CYCLES, photo-elicitation has proved to be a successful method of generating talk and insight around consumption objects and practices (e.g. Cappello, [5]; Cody, [7]; Epstein, Stevens, McKeever, & Baruchel, [11]) and about experiences of living within particular neighbourhoods (Dodman, [10]; Holloway & Valentine, [21]; Lomax, [26]). Using participants' images in interviews allows them to 'show' places or areas of life that may otherwise be inaccessible to researchers (Croghan, Griffin, Hunter, & Phoenix, [9]). Our instructions to photograph specific dimensions of everyday consumption aimed to include items, practices, places and issues which may not have been remembered or referred to if participants were simply asked to discuss the topic.

This does not mean, however, that we assumed that through photographs we would gain access to more 'authentic' accounts of participants' everyday lives. Photography is necessarily selective and research participants' images are shaped by 'what is photographable' (Barker & Smith, [2]), the representations of self they want to communicate and the guidelines given by researchers. We did not employ photo elicitation so much as a means to gain access to the 'reality' of everyday consumption as to enable richer group discussion. It is not that interviews in which photographs are discussed are necessarily 'better' (Lomax, [26]), but that photographs bring something different to the encounter (Croghan et al., [9]; Harper, [18]).

An additional motivation was the potential which young people's photographs offered for illustrating and communicating the project (Rose, [38]) and generating public engagement with the work – both within each city and internationally. As Luttrell notes, photography can work to 'rouse and reframe conversations' ([27], p. 225) between researchers and participants, amongst research participants and among viewers/readers (see also Puwar & Sharma, [34]). From the start we planned to exhibit participants' photographs and incoporate them into reports and publications.

An outline research design for the photo elicitation task was drawn up by Burningham, Venn, Nissen and Hayward after detailed discussion with all project partners during a face to face meeting. The design involved initial focus groups (with 6–8 young people in each age group of 12–14, 15–17, 18–24), a photo task and then a second focus group in which participants' images were used as the basis for discussion.

The initial focus group explained the project rationale and explored participants' everyday consumption routines. Participants were then asked to create a photographic record of a day in their lives during the following week, providing images of: their everyday mobility (how they get around); their 'normal' home-life (of energy and water use); their food consumption (what they eat, where and with whom); their most common and enjoyable form of leisure (how they relax and communicate); and something that represents how they spend most of their day (e.g. employment or education). In addition all participants were asked to present images that showed something they like most about their local community and something they would like to change. Follow up focus groups used participants' photographs to elicit discussion about everyday consumption and broader reflections on what aspects of local life enhance or constrain their ability to live well. All focus groups were recorded and transcribed in full. Dissemination plans included an illustrated report and exhibition.

Overarching ethical guidelines for CYCLES were developed from the UNICEF guide for Ethical Research Involving Children (ERIC) (Graham et al., [15]), while acknowledging that each city team would have diverse cultural and institutional ethical requirements. The motivating principle behind ERIC is that the human dignity of children and young people is honoured, and that their rights and well-being are respected in all research. Each partner completed a form providing information about how they would seek to minimise potential negative consequences of participation in CYCLES and confirming their commitment to the principles of informed consent, confidentiality and secure data management. All agreed initially that: participants' contributions would be anonymous; pseudonyms would be used in all transcripts and reports; and participants would be encouraged not to take photos of people's faces. In addition to completing this form, researchers in the UK, NZ, Bangladesh, Japan and South Africa complied with their University or Research Institute's requirements for ethical review. All photographs and anonymised transcripts were shared through a secure cloud storage service hosted by the University of Surrey and backed up regularly to secure servers. Access to this storage service was managed through a designated login and password for each research team who were given their own secure partition which was inaccessible to other users.

The outline research design and ethical guidelines provided guidance for partners, but we anticipated some variability in how the research would actually proceed in each country. In the following sections we reflect on some of the issues that emerged in practice, first concerning technologies for taking and sharing photos and then around the content and use of participants' images.

The development and spread of digital photography means that young people are increasingly experienced photographers. The simplest way for research participants to produce good photographs and to share them with the researcher is to take them with a smart phone. However this immediately raises issues of whether all participants have access to a smart phone, whether the use of phones is considered age and culturally appropriate and how images will be securely shared with the researcher. Our commitment to culturally sensitive ethics in context meant that the appropriateness and acceptability of photo methods had to be considered on a case by case basis rather than imposing standard requirements across all of the countries.

Disposable cameras have commonly been used in research with children and young people (e.g. Barker & Smith, [2]; Rasmussen, [36]), but the increasing ubiquity of mobile phones means that children are now quite unlikely ever to have used one. The 'strangeness' of disposable cameras has rarely been reflected upon in literature about photo elicitation, but was evident in our project.

Disposable cameras were used in the UK and NZ where it was recognised that not all younger participants would have access to a smart phone, and/or their parents or guardians may have concerns about them using one for this task. In these countries the method for taking photographs was age specific; in NZ, infomed by parental concerns (see discussion below) all under 17s used disposable cameras, while in the UK it was planned that those under 15 would use disposable cameras (as regulations restricted the use of WhatsApp which we planned to use for sharing photos to those over 15 (see below) and those aged 15–17 were to be offered a choice of using a disposable camera or their phone).

Disposable cameras, and instructions on how to use and return them so that the photographs could be developed, were given to participants at the end of the first focus group. Before the second focus group started, each participant was given a full set of the photos they had taken and was asked to tell the researcher if there were any which they did not want to be included in the discussion. At the end of the group participants were asked again whether they wanted to remove any photos from the research and also invited to identify any which they particularly wanted to be exhibited. They were all given a full set of their own photographs to keep.

When participants were given disposable cameras they responded with excitement and puzzlement. In NZ one of the younger participants was so fascinated by the disposable camera that the interviewers gave him a spare one to take apart rather than him dismantling the camera he was supposed to use. In the UK each participant was given a stamped envelope addressed to the researcher, and asked to post their camera back once they had taken all their shots. The stamped envelope also provoked surprise, with some participants never having had one to fill and post before. The researcher explained how to post back the camera so that the film could be developed and the photos ready for discussion at the second focus group. However some participants clearly did not understand this – one sent back the stamped address envelope empty except for the instructions on how to use the camera, and another arrived at the second focus group with the disposable camera in expectation that the photos could be developed there. We had elected to use disposable cameras for good reasons but had not anticipated how they would be received by participants. In practice, the strangeness of the technology brought an unexepected element of fun and interest into the research task for most of the participants. This highlights for us that what young people actually find engaging about visual methods may turn out to be something quite different from what researchers anticipate.

In Bangladesh, parents of younger participants did not want them to use cameras, so the Bangladeshi researcher asked them to draw pictures instead. Using drawings to elicit interview discussions is a well established method in research involving children (see Angell, Alexander, & Hunt, [1]; Mitchell, [32]). Pictures were drawn at home and then brought in to the focus groups and discussed. Few of the older participants had mobile phones. In some cases this was an issue of affordability, but for others it related to parental concerns about the use of digital devices, in particular that they would be a distraction from studying. Participants from wealthier families in Dhaka used their parents' phones to take photos, enabling their parents to check the images and be reassured about what the phone was being used for. Those from lower-income families were usually able to borrow phones from other relatives or friends' parents. The principle of older participants independently taking photos which would be shared securely with the researcher (a fundamental part of the ethics application for the project in the UK) proved to be inappropriate and impossible in Bangladesh, where parents often wanted to scrutinise the pictures being shared with the researcher.

Where children use cameras away from the researcher, other family members may step in and shape what is photographed to provide a particular view of cultural or family life (for example, see Mannay, [29]). In cases where children are using a relative's phone there is an even stronger possibility that some of the photographs may not have been taken by the child, or at least that the choice of what to photograph may not be entirely independent. This became evident in a focus group in Japan when younger participants were discussing the images. One of the photos was a particularly striking image of food (Image 1) ; the participant who had submitted this photo commented:

'This is quite different from my ordinary breakfast. My mum made this breakfast to be shown to others!'

PHOTO (COLOR): Image 2. Inclusion of non-identifiable people in photos

This instance reinforces the need to treat claims that visual research 'shows' the 'reality' of young people's everyday life with caution. However in the context of the focus group it provided the basis for a discussion about what 'an ordinary breakfast' was – thus the image worked well to facilitate discussion about everyday consumption despite not depicting it. The incident also contributed to our critical reflection on issues of potential shame and pride in international visual research which we discuss later.

In India younger participants did not have access to phones and the collaborative production of images for the project became an explicit part of the research design. The data collection in India was conducted by Swechha team members, with participants recruited from those they worked with in Jagdamba Camp (a slum community in New Delhi). The project was thus situated within and shaped by Swechha's commitment to intensive interaction with young people to engage them in the process of change. Rather than taking photos individually, researchers went with a group of younger participants to their neighbourhoods and homes. The children discussed what to photograph and then took photos using the researchers' phones. In some cases where children were not happy with the quality of the image, Swechha team members took photos under the children's direction. The taking of the photos thus became a group endeavour and the opportunity for discussion about what the children wanted to show of their community.

Our discussion here highlights that what we initially considered to be primarily a practical issue of how to provide access to cameras led to a range of unexpected learning. Using disposable camers in Lambeth and Christchurch revealed how technology can enhance participants' engagement with research in unexpected ways. The fact that across the project individuals did not all have access to their own means of taking photos accentuated the extent of collaboration in the production of visual images – not just between participants and researchers but also within groups of participants and between participants and other family members. The particular concern evident amongst parents in Bangladesh, and NGO staff in India, about who took photographs and what images could be shared, can be understood in the context of communities in the global south increasingly exerting their right for control over local images in response to a history of journalists taking exploitative and clichéd photographs (Idris, [22]). Recognition that images were co-produced challenges ideas that visual methods necessarily provide access to individuals' own representations of their lives, but provides insights into community and family relations and concerns (Mannay, [29]). Our discussion of how the issue of access to cameras was dealt with in some of the countries emphasises the need for flexibility of methods and approaches to ethics in international visual research. It was not only that providing children with individual access to the means to take photos proved practically challenging in some of the research contexts, more importantly it would not have been socially acceptable.

WhatsApp was initially considered as a platform for sharing electronic photos because of its end-to-end encryption which protects communications between two users so that only they can read the messages. In this section, we reflect on some of the issues which arose around the use of WhatsApp.

Social media apps offer researchers using visual methods with young people new and potentially convenient ways to share images. We envisaged that researchers would set up one to one Whatsapp groups with each participant to ensure their contact details were not shared with other participants. However, in South Africa participants over 18 years were keen to share their photos and thoughts on them in a WhatsApp group with other participants, and the researchers decided to enable this. Methodologically this worked well to stimulate on-line comments clearly linked to specific images, but also fostered a sense of group identity, ongoing discussion and collective desire for action amongst participants. Thus, young people in Makhanda used WhatsApp as a way to take control of the process of sharing and discussing their images. For these participants the desire for ownership and control of their images trumped concerns about anonymity and privacy.

However, concerns about the potential misuse of the app – either by adults conversing with young people, or by young people themselves – meant that its use was not appropriate everywhere. Within the NZ context using a secure social media app in this way turned out to be a source of perceived risk. At the time there was national media coverage of a politician's use of Snapchat to message young people, with attention not only focusing on the individual case, but on broader isues surrounding ethics and the use of social media apps (e.g. NZ Herald, [33]; Stevens, [40]). This created an environment in NZ of sensitivity about adults engaging in one to one social media interaction with young people, precisely because of the extent of privacy guaranteed by the use of the apps. In light of this, the NZ team decided that rather than using WhatsApp, participants would share their electronic photos with them by saving them onto a memory stick. This incident illustrates how quickly cultural evaluations of appropriate ethical practice can shift, with assesments of the 'safety' of technology only ever being provisional.

In the UK, researchers planned to offer participants over 15 the opportunity to use WhatsApp to share photos securely with them. However, during the project WhatsApp changed their regulations, restricting the app's use to over 16s (Gibbs, [13]), meaning that plans had to be changed to ensure that 15 year olds used disposable cameras. The constantly evolving social media landscape thus proved to be another element which contributed to the need for ongoing flexibility and responsiveness in our research design.

Our ethics guidelines included the clauses 'participants will be encouraged not to take photos of people's faces' and 'participants' contributions will be anonymous'. Such commitments are common in visual research, although critical perspectives on the implications of both are increasingly voiced (Wiles et al., [44]). In this section we reflect on how these commitments were responded to by participants and researchers across CYCLES.

In Japan, NZ and the UK the research teams ensured that no photographs including identifiable people or locations were used. For participants used to taking selfies this increased the novelty – and perhaps the difficulty – of the research task. The project required them to look out from themselves, resulting in distinctive images of homes, parks and streets often without any people in them. Participants reflected on the strangeness of taking pictures of mundane aspects of their lives – such as the bus they got to school – rather than of themselves or their friends. However some of the older participants reflected on how the task had helped them reflect on their neighbourhood and the good things about it, suggesting that the project at least partly realised the potential of encouraging young people's engagement with their local environment – as one participant in Lambeth commented 'when you do take photos of things you realise what you do love about the place.'

Older participants in Christchurch, New Zealand reflected on how the requirement not to photograph people had led to dimensions of social difference and inequality being under-represented. They noted that 'in the city ... there's heaps of homeless people ... and you weren't allowed to take a photo of that and you can't really see that' and 'if you were allowed to take photos of other people you'd see that it's quite multicultural'. While the experience of social inequality undermines the potential for sustainable lives (Wilkinson & Pickett, [46]) and good social relationships are fundamental for young people's sense of wellbeing, our ethical concerns with informed consent and the protection of privacy inadvertently prevented young people from showing these things in their images. Participants' discussion of these visual absences underlines the importance of contextualising the images in the light of the subsequent discussion, and reminds us that visual research should not just be interested in what is shown, but also what is not (Mannay, [29]).

While conforming to the requirement not to include identifiable people in their photos, participants were often creative in the ways in which they ensured that they and other people were present in some of their images. We had multiple images of participants' feet, of hands on bicycle handlebars and of friends photographed from behind (Image 2). These images placed participants within the contexts they were showing and discussing and illustrated the importance of others within their urban landscapes. For instance, one particularly evocative photograph depicted a participant's foot in her school shoe, alongside that of her friend standing on yellow lines painted on the street – without showing any identifiable people the image speaks loudly of the importance for young people of doing things with friends in the city.

PHOTO (COLOR): Image 1. Photo of breakfast made by Japanese participant's mother

In Bangladesh, Brazil, India and South Africa the commitment not to include identifiable images of people was challenged as the project progressed. In India the organisation of photography as a group activity and opportunity for discussion about what was valuable within their neighbourhoods resulted in strong messages from participants that they wanted to show people. Similarly in Brazil and South Africa, older participants argued that images of their families and friends were most important in conveying the aspects of everyday life which they valued. In all of these contexts the researchers sought to ensure that anyone shown in a photograph had consented to the inclusion of the image. Most of these images show friends and family members smiling and spending time together, providing positive images of the significance for young people of their relationships.

For participants over 18 in South Africa, including themselves and their friends in photographs was part of a broader rejection of the conventions of anonymity in visual research. They argued strongly to be recognised as the creators of their images which some described as 'art work', seeing the demands for anonymization and removal of identifiable people as problematic appropriation (Wiles et al., [44]). The South African researchers decided that it was important to respect participants' wishes. After careful checking of institutional permissions and consent processes, images of some participants over 18 were included and their real names were used.

In Dhaka the inclusion of images showing identifiable people was the topic of particular discussion. Transport is a major problem within the city; participants in focus groups described the severity of congestion and how inadequate and unsafe public transport was. Reporting incidents or accidents to the police was often considered pointless, with young women in particular fearing police harassment. During the course of our research young people spontaneously took to the streets to protest about road safety and traffic management (British Broadcasting Company (BBC), [4]). Our participants captured the protest, including photographs of children holding banners with slogans (in Bengali) such as 'is it a crime to ask for justice?'. Being involved in these protests was dangerous, with potential repercussions for those identified by the police. Participants in Dhaka were keen for their images of the protest to be shared, so that an international audience could see young people's strength of feeling about the issue. After ongoing discussions about how to respect their wishes, but protect those pictured in the context of the longevity of images and uncertaintly about how they might be used (see Brady & Brown, [3]; Mannay, [30]), it was decided to include the images within the temporary exhibition in London, but not to include them within any publications. As Wiles et al. ([44]) reflect, deciding what visual images to show and how, is not simply a matter of respecting participants' wishes, but also involves researchers carefully reflecting on potential harms which may arise from dissemination.

The fact that some of the teams decided to include identifiable people in images meant that there was some unevenness of images in our international report and exhibition. Observers might assume that the lives of young people in Makhanda, New Delhi and São Paulo are more social and their relationships more central than is the case for young people in Lambeth, Christchurch or Yokohama – while in reality what was included or excluded from the images was a result of the approach to anonymity adopted in each context.

How to navigate ethical concerns around anonymity while acknowledging participants' desire to be recognised is a longstanding debate within visual research (Wiles et al., [44]). The examples above illustrate how the multiple cultural contexts and numerous researchers involved in our project made issues around sharing photographs which included identifiable people even more complex. Consent was often multi-level, being gained informally from those photographed by research participants, formally from research participants by researchers and by the UK and NZ project leads from the researchers within each country. Researchers' concerns sometime overrode participants' consent, with partners requesting that some images be removed from the shared project files (see Mannay, [30]). Just as participants' informed consent is an ongoing process, so too was consent from research partners for the use of images from within their countries. As we compiled materials for the project exhibition and booklet we constantly checked with partners if they were happy with the use of particular images and removed any which they had concerns about.

In the qualitative components of international projects, comparison typically relies on a case oriented approach where the focus 'is on differentiation and diversity between and within countries, societies and cultures, and the complexity of the factors involved' (Hantrais, [17], p. 135). In CYCLES, visual research provided the potential for identifying striking comparisons – and similarities – between the contexts and experiences of young lives in the seven cities we were researching. However, this potential was also a particular cause for concern for research participants and researchers when it came to selecting images to be used in the exhibition and accompanying report.

Participants in several of the cities expressed unease about how their images might depict aspects of their everyday lives unfavourably in comparison with images from other countries. This was a specific concern for those from poorer countries – and from poor families in rich countries – who did not want to highlight what was lacking in their lives. This was particularly evident in relation to images of food; the lack of adequate food was not photographed although this was a reality of some participants' lives. Even for richer participants, concerns for their diets to be portrayed favourably were evident. This was dramatically illustrated in the Japanese example discussed earlier, in which a participant noted that her mother had set up a photograph of a lavish breakfast to look good in international comparison.

Concerns to avoid participants' feeling shame about their lives underpinned the decision by the Indian researchers to re-orient the research in New Delhi to focus on what young people wanted to celebrate and share internationally. However selecting photographs for display also raised wider concerns for researchers working in decolonizing contexts, in South Africa, New Zealand and India, about potential power imbalances in curation which could result in 'othering', exoticisation and romanticisation (Lutz & Collins, [28]; Smith, [39]).

For the Bangladeshi research partner there were particular concerns about how the children's drawings would be viewed in international context. While children's drawing are easily viewed as 'cute', he was at pains to point out the circumstances of their production and discussion (Mitchell, [32]). When considering whether to exhibit a child's drawing of a rickshaw, the team was conscious that some children in Dhaka had been beaten and imprisoned for protesting about transport, and therefore inclusion of the image was not about 'a nice drawing of a rickshaw'. Concerns arose partly because of the dual way in which we were using the images both for eliciting discussion within focus groups and for international exhibition and project dissemination. While sensitive contextualised discussion of young people's photos or drawings could be facilitated within the focus groups in each city, researchers were aware that they could not easily control how the images were viewed once they became part of an international exhibition and publications.

In fact the very concept of an 'exhibition' of the images in London proved to have unwelcome echoes of colonial histories in which artefacts from colonised countries were taken and exhibited in the UK. We chose the venue carefully so that it was in line with project aims and ethos, opting for The Foundry, a centre which offers office, meeting, conference and exhibition space to social justice and human rights focused organisations. Careful ongoing conversations with partners were necessary to build mutual trust, with details of the venue and drafts of the exhibition materials and booklet being sent to them for approval. Images were grouped into broad themes (e.g. My place, my space; My city, my home; My everyday food; Getting around in our cities) each of which was introduced by a board contextualising them with an overview of discussions in the focus groups. Individual images were accompanied by quotes from participants and subtly colour coded to enable viewers to identify which city the image and quote was from without making comparison between the cities the dominant focus.

Research participants could not be involved directly in the production of the exhibition and report because of their geographical spread, and the tight time frame we had to produce outputs. However their choices of images to include were communicated to us via the national researchers and we sought to include their voices through quoting their own words. In practice the co-production of the outputs was between the large international research team, with negotiation about what was appropriate ongoing right until the exhibition opened and the report went to press.

The exhibition was billed as an international touring exhibition, with the intention being that all partners could exhibit the images in their own cities. We were able to provide funds for all the partners – and some of the researchers they worked with – to attend the launch of the exhibition in London. Researchers from Japan, India, Bangladesh, Brazil and South Africa spoke as part of a panel session at the event to ensure that they were able to contextualise the images from their own cities.

Existing literatures have discussed both ethical issues raised in visual research with young people, and the problems associated with applying 'universal' ethical guidelines across cultural contexts. There has been little consideration to date, however, of specific issues raised in projects where visual research is being conducted with young people simultaneously in multiple national contexts. In this paper we have shared some of our experiences in CYCLES in order to contribute to knowledge in this area.

Throughout we have reflected on the multiple ways in which we had to adapt our expectations around the way the research would proceed, illustrating how technical, methodological and ethical considerations are often intertwined in practice. Flexibility was necessitated: by unequal resources and access to technology across the research teams; issues of cultural acceptability; shifting social media landscapes, and participants' and researchers' demands for greater influence on the production and content of images. We identify strongly with Gallacher & Gallagher's ([12]) sense of often 'muddling through', and embrace their recommendation to adopt an attitude of 'methodological immaturity' in our research. This positioning 'privileges open ended process over predefined technique' ([12], p. 513) accepting 'social research – and life more generally – as a necessarily complex, incomplete and messy process' ([12], p. 511). It is an approach which accepts the need for research design and ethical approaches to evolve and change within projects and emphasises the need for an attitude of flexibility, openness and learning.

The need for ongoing negotiations around research design and ethics in international comparative research brings the benefit of researchers gaining increased understanding of the other cultures with which they are working, and critical reflection on their own research processes (Hantrais, [16]). In CYCLES the necessity for flexibility turned out to provide a range of unexpected additional learning. What initially seemed to be a practical issue of how to manage access to technology to take and share photos ended up being a process through which we learnt about unexpected ways in which technology might enhance participants' engagement; about the co-production of images and how young people are situated in family and community relationships. We also developed a more critical perspective on the concept of privacy. While privacy is enshrined in institutional codes of ethics, in practice it is not always locally appropriate or acceptable. In New Zealand the privacy of communication assured by social media apps emerged as a source of concern; Bangladeshi parents wanted to check the images their children were sharing with researchers and in South Africa participants rejected anonymity and private communication with the researcher in favour of building dialogue and relationships with each other.

A longstanding challenge in international research is ensuring sufficient coherence across a project to enable comparison, while remaining committed to the flexibility demanded by an ethics in context approach. As our visual research progressed, however, it became clear that potential international comparison between young people's lives in different cities was itself politically and personally sensitive. Concern about the possibility of unfavourable comparison informed both the images participants produced and the way in which we exhibited them. At this stage of the project we were able to exhibit images to bring out particular issues for young people in each city, alongside common priorities, without attempting systematic comparison between cases. How to approach the issue of international comparison is an issue we will remain attuned to as we move into the subsequent research phases of CYCLES.

Building relationships of trust and being committed to ongoing negotiation within the team have been integral to enabling CYCLES to progress. This is not easy given multiple partners and time zones and unreliable internet connections, as well as the sensitivity of many of the issues raised. However through the process we have gained greater understanding of each other's positions and of the contexts for young lives in the cities we are studying. The concept of methodological immaturity was developed specifically for childhood research, but we suggest that it is also particularly valuable in international visual research, where flexibility, humility and willingness to learn are all essential. This approach is increasingly hard to maintain in a research context governed by ever tighter ethical and data management requirements, but it is essential in navigating the complexity of international visual research with young people.

special thanks to the research assistants in each country: Ashraf Uddin Fahim; Sofia Ferraz; Aakriti Gupta; Md. Tanzid Hossain; Tanvee Kakati; Kendall Lattin; Luisa Leo; Mandilive Matiwane; Lona Musiyiwa; Patience Shawarira; Nuku Tau; Niwa Ututaonga; also to Linda Gessner for her expert assistance in producing the visual images; to Susan Bouterey and Zea Harman for translation and to all our research participants. CYCLES is funded by the ESRC as part of CUSP (The Centre for Understanding Sustainable Prosperity) ES/M010163/1.

No potential conflict of interest was reported by the authors.