Chapter 5: Ethical Considerations During Photo-eliciting Trajectories with Migrantised Women Focused on ‘Gender Empowerment’ in Civil Society Organisations

Visual methods, such as photo-voicing and photo-eliciting methods, are becoming increasingly popular in social sciences. These visual techniques are used to conduct research when working with participants who are stigmatised, vulnerable, oppressed, or who find it difficult to articulate ideas solely in a narrative or interview format (Clark, 2019; Milne & Muir, 2019; Teti, 2019; Wang, 1998; Wang & Burris, 1997). Photo-voicing can be used to discuss and conceptualise difficult questions and to express ideas beyond language (Teti, 2019). The value of using such methods lies in their high level of flexibility and in the possibility of adapting them for use with particular target groups and communities (Wang & Burris, 1997). In doing so, researchers aim to generate insights that they would not have access to through other approaches, and likewise participants can raise aspects the researcher would not have thought of as relevant (Cooper & Yarbrough, 2010; Wang & Burris, 1997). This facilitates the discussion of controversial topics or topics where the researcher and the interlocutor have diverging takes on what is being discussed (Wang & Burris, 1997).

Despite the growing popularity of visual methods (Hannes & Parylo, 2014; Milne & Muir, 2019; Rosemberg & Evans-Agnew, 2020; Teti, 2019), less attention has been given to the ethical challenges and questions related to these visual methods (Clark, 2019). These difficulties concern the uniqueness of visual methods, which complicate possibilities for maintaining anonymity and confidentiality, jeopardising the approval of ethical boards at universities or funding bodies. This chapter reflects on the ethical considerations that we encountered when working with visual research methods in our research focused on migrantised groups.¹ This is relevant since, in contrast to audiovisual/narrative techniques, some additional questions arise. These issues range from safeguarding the participants despite their higher visibility through photo-voicing, exhibiting photos, problems about ownership of these images, the potential for photos or images provided by the participants to be misinterpreted, as well as issues of anonymity and pseudonymisation. In exploring these issues, we support the idea that ethics are not set in stone but constantly evolve over time and are adjustable to the context (Liebenberg, 2018).

In the following sections, we provide an overview of photo-voicing and photo-eliciting methods and consider the existing literature on ethics related to visual methods in social sciences. Afterwards, we reflect on the ethical issues and considerations related to visual methods made throughout the ReIncluGen project.

Since Collier and Collier (1967) introduced photography as a research method, coining the concept of visual anthropology, researchers have used, experimented and reflected upon how visual methods – in particular photographs – can go hand in hand with research. There are several visual methods, ranging from journey mapping, photo-voicing and photo-eliciting, used in social research. While both photo-elicitation and photo-voicing involve the use of photographs to gather qualitative data, the key difference lies in who selects and takes the photographs.

Photo-voicing involves participants taking their own photographs to share their perspectives. The core of the photo-voicing methodology is that the studied group receives photography training and is instructed to take pictures of anything they deem important regarding a specific subject. Afterwards, the photos are shared and discussed individually or in a group in the format of a focus group discussion. Photo-voicing can be used as a technique, primarily focused on the pictures, or as an auxiliary technique to assist with interviews on complicated topics (Duffy, 2011; Wang & Burris, 1997). Wang and Burris (1997) argue that the use of photo-voicing techniques enables participants to (1) amplify their voice, (2) bring to the fore less acknowledged types of knowledge, such as tacit and experiential knowledge, (3) be a catalyst for social change and/or be a tool for the development of critical consciousness, and (4) change existing dynamics in relationships between communities, researchers and policymakers. However, Milne and Muir (2019) have added some critical perspectives, highlighting the pitfalls that such methods may encounter and questioning whether photo-voicing as such will necessarily result in such positive social changes. Rather, they suggest, these could lead to unanticipated effects, including suspicion of the researchers, and negative portrayal of the communities that participants belong to.

Photo-elicitation involves researchers providing photographs to participants. Photo-eliciting techniques are defined as follows by Harper (2002, p. 1): ‘Photo elicitation is based on the simple idea of inserting a photograph into a research interview’. According to Harper, the use of images, such as photos, drawings and cartoons, helps to evoke more distinct kinds of information than would have been retrieved using interviews. This technique heavily relies on the idea that visual information enables the voices of participants to be amplified (Luttrell, 2010; Milne & Muir, 2019).

In our project, we aimed to consider the diversity among our participants. This influenced the choice of one method over another. We focused on the ethical considerations when using visual research techniques. In doing so, we sought to create an environment in which participants use photos or objects, which could be older or more recent ones, taken or chosen by participants to denote their understanding of ‘empowerment’ in general and to further specify their thinking about gender in this context. Using and selecting these typically led in two different directions: (1) the use of photographs or objects to gauge meanings of empowerment in consecutive interviews, and (2) to voice the views and experiences of participants in relation to empowerment to a larger audience in a photo-exhibition, accompanied by a small text or story told by the participants on the significance and symbolisation of the photograph. This supported the viewers of the photos to better interpret or (re)imagine the meaning of the photograph. Participants were first asked to participate in the interview and afterwards to give permission to share their photographs and narratives in the exhibition.

When reflecting on ethical considerations regarding the usage of visual methods, it is important to distinguish between the ethical concerns that arise across diverse target groups and research settings. The formulation of the questions to be asked about images or the demand to bring along or take pictures also requires some cultural sensitivity.

Many studies using visual methods focus on vulnerable groups in society, including children (Epstein et al., 2006; Glaw et al., 2017; Gube, 2022), which need additional care and attention. For instance, in a study by Epstein et al. (2006), in which photos were used to elicit children’s perspectives, the authors highlight that some ethical concerns arose in relation to who takes the picture (whether researchers or the children) and the extent in which they are involved in research. In that study, children who were attending a camp were given a camera at particular moments in time (after the camp activities), so they would not always be in charge. In another study, by Glaw et al. (2017), the researchers wanted to compare photos taken by those who suffer from mental health complaints, with those taken by a group of scholars. However, as ‘people in an acute care setting may be very ill and may struggle to participate in research of this type’, they opted to select solely people in a community setting (Glaw et al., 2017, p. 4).

A second theme is that visual methods could be of added value, especially in specific, often sensitive, settings (Gautam et al., 2018). This may be so where target groups are less easy to reach through other research inquiries, such as quantitative surveys. For example, Gautam et al. (2018) conducted research with sex-trafficking survivors in Nepal. Based on their experiences, they suggested using photovoice during rehabilitation.

Third, some projects, like the SEXHUM project led by Nicola Mai, seek to ensure that more efficient and ethical policies and social interventions are developed to address the needs of the research participants, in this case sex workers in strategic urban settings in Australia, France and the United States.² The methodology for this research draws upon several qualitative methods like art-science ethnographic filmmaking in which research participants partake in the ethnographic filmmaking to narrate their realities and to address the relational and performative dimensions of their experiences.

Finally, it needs to be clearly defined who will eventually see these pictures after the research has been conducted. These decisions will necessitate clear communication and discussion with the participants – and sometimes an additional ethics review may be needed to clarify these boundaries.

In our project, which is still in progress, photo-eliciting methods were deployed to help us understand the concept of ‘gender empowerment’ by migrantised women in an emic, inductive and collaborative way. Asking women to bring pictures and using them as an entry point allowed the interview to be led by the ideas and concepts of the women, rather than to have preconceived questions and indicators of gender empowerment imposed from outside by the researcher. In narrative interviews, the questions, their formulation and the order of how they are asked can predefine the way interlocutors formulate their answers. By using photo-elicitation, we wanted to avoid this and let conceptualisations of gender empowerment stem from the participants and their views. In this way, it was the participants who led, decided and selected the topics, the approaches and emphases. This enabled us as researchers to understand more deeply why the photos were chosen or taken as they were, what was being expressed and why this was important to them and the group. Although we initially focused on photographs during the preparations for our fieldwork, and during the interviews themselves, we realised that not everyone is familiar or comfortable with taking photographs, and that they could feel burdened to capture a sensitive and broad concept in a specific photograph. Because of this, we opened up our methodology so that participants could also, if they wished, bring objects to the interviews.

By directly including participants in several steps of the research process, this research project aimed to be highly co-creative (Liebenberg, 2018). This participatory dimension can be seen first of all in the inclusion of civil society organisations (CSOs) within our project consortium and second by our use of participatory research methods, such as photo-eliciting interviews resulting in an art exhibition. Furthermore, our commitment to participatory approaches was reflected in our use of evaluation methods to assess promising practices of gender empowerment and inclusion in CSOs using an intersectional approach. As such, we found that these approaches could be empowering, shifting some of the control over the research process and the main categories and definitions from the researcher to the participants (Budig et al., 2018).

Through the creative process of selecting pictures and attributing meanings to them, participants took the lead in defining and crafting meanings, which allowed for inductive and collaborative research. Hence, the selected or produced picture, combined with an explanatory quote or essay, could also result in more ownership, especially when making them public (Wang & Burris, 1997). As visual methods are exploratory, inductive methods, we aimed to deconstruct the meaning of the concept ‘gender empowerment’ and to stimulate debates about what this concept could mean, how it could be (re)interpreted in society, and how it is applied in the CSOs that participants are involved with.

During this project, for each of our case study countries (Austria, Belgium, Italy, Spain and Poland), we selected three CSOs and invited 10 migrantised women who participated in the CSOs’ activities to conduct approximately three consecutive interviews (30 participants in five countries: 150 participants in total). We realised this goal, although some participants preferred only to have a brief contact or dropped out, especially for the third round of interviews. Before the interviews, the team conducted around 20 days of ethnographic fieldwork in each of the CSOs. This helped to build up trust between the researchers and the participants and to encourage participants to engage in the three consecutive interviews. During the first interview encounter, ethical aspects were explained and discussed, and the researcher asked about their initial ideas on the concept of ‘empowerment’, and especially ‘gender empowerment’, beginning to build a rapport with the participant and then explaining the ‘photo assignment’ and discussing together what the photos should depict. This is a crucial step in encouraging participants to share photos, as many are hesitant or feel inadequate at first, and to discuss how a relatively vague concept such as ‘gender empowerment’ could be portrayed in photos.

Participants were then asked to bring around three to five photos to the next interview, which we hoped would be less stressful than having to select only one. Some prompts were also given to ease the respondents who felt inadequate to use pictures to tell their views (for example, ‘bring a picture of a moment where you feel strong, encouraged or like yourself’). It was, however, emphasised that these guidelines were only suggestions and that they were free to bring whatever felt right. In this sense, participants could also bring a personal or symbolic object. In the third interview, participants were asked about their opinion on photo-elicitation how they felt after having reflected in such a manner on their processes of empowerment and whether they would still like to add something to their narratives. Additionally, participants were asked whether they would be willing to showcase one of their photos or objects, initially used during the interviews, in a travelling art exhibition and photo-essay book, in which they would present a short explanatory essay or quote to set out their ideas concerning these photos and their message. At this point, they would be asked to select only one photo together with the researcher. Although the interview data were processed in an anonymous way for analysis, for the photo-exhibition, in contrast, participants are asked whether they want to add their name to the photo or prefer to use a pseudonym, and if they would like additional changes to be made for reasons of privacy, such as blurring parts of the photo. From our perspective, the choice to see all participants at consecutive moments over time, even after extensive ethnographic fieldwork, was needed to develop trust. This reflects the importance of helping people to feel at ease before being involved in ‘creative’ activities of this kind, which is also mentioned by Shobiye (2023), who asked refugee women to make simple drawings at the end of the interview on mothers’ social spaces.

Next to the photo-voicing interviews with women who participated as part of the collaborating CSOs, we also launched a photo-essay contest in order to widen our scope to a broader public and reach other migrantised women in society. Through their participation in the contest, other women’s photos and objects could also become part of the international travelling exhibition. Another rationale behind the contest was to raise awareness and to promote our project with society as we diffused the contest call using different channels, including via social media, newsletters, emails, project and CSO websites. In the following section, we discuss the four main themes of our ethical reflections during the fieldwork.

As explored below, undertaking this research involved us in negotiating ethics before, during, and after the photo-eliciting trajectories within CSOs and with their participants.

Before we started the fieldwork, all university partners asked for approval from their research ethics committees (RECs). We found that doing this took quite some time and so slowed down the work of reflecting on ethical issues as a team. Given that the research funder had also required the appointment of an Independent Ethics Advisor to provide advice, support and monitoring of the ethical issues arising in the project, the appointed Advisor also provided feedback on reports to the funder as well as on the submissions to the local ethics committees.³ In view of the detailed attention paid to these processes, this all took quite some time.

Some reflections can be made that may be useful for ethical boards at research institutions or for university networks. First, before submitting, the researchers were obliged to discuss their proposals with an ethics officer and a data manager. These discussions led to the joint development of a ‘decision tree’ on anonymisation and pseudonymisation (below) presenting ethics more as an ongoing discussion that needs to be continuously considered and adjusted to the actual needs of researchers and research participants, as well as to the changing requirements and developments in society and the research project itself. Similarly, the requirement to discuss ethics with an external Independent Ethics Advisor obliged us as a team to discuss our ethical concerns in some depth. This resulted in further reflections on trauma-informed research competencies, highlighting the need, from our point of view, for ethics committees to take account of these issues (Goodwin & Tiderington, 2022). This caused the researchers to organise a lecture on this topic by Goodwin and Tiderington (2022). These are good examples of seeing ethics in terms of an ongoing discussion about issues that are not ‘set in stone’, a perspective that we feel is not sufficiently adopted in academia, where there may not be much space for these debates.

Second, we found reflecting on ethics and submitting protocols to the ethics boards to be a time-consuming process, with diverging speeds and demands across universities that could be simplified by collaborative ethics networks, in which the same ethical requirements were shared per university, avoiding duplicating work for ethical boards. As these steps are time-consuming, we would argue that (1) protocols and related documents such as informed consent forms (ICFs) should be evaluated as swiftly as possible by ethics committees, consistent with their duty to scrutinise these carefully; (2) where appropriate, requirements for ethics reports across research institutions should be standardised, and a clear timeline for acknowledgements of these reports should be introduced; and (3) a processual approach to ethics should be included, in which at several moments in time, researchers have the opportunity to discuss and evaluate their ethical considerations with an ethics officer.

Third, collaboration and unification of participant information across research institutions may yield benefits. During the course of the project, a standardised form in ‘simple language’ was developed. Especially for participants who are not aware of these ethics procedures, who do not trust governmental bodies, do not speak the local language and are suspicious of signing formal documents, the language and content of ICFs can be burdensome. The use of simplified ICFs can be helpful here. Collaboration across ethics boards could further facilitate the development of ICF templates in different languages, to which only small changes may need to be added regarding the content and objective of the research. This can facilitate awareness of ethics and makes sure all participants are really informed about the ethical procedures followed.

Of the three CSOs involved in our research, one is actually part of the research consortium and the other two were selected later on. The process of negotiating access depended on whether or not they were included in the consortium and their size, which had an impact on how many people in the CSO were already aware of the research and its objectives. In addition, some people working with the CSOs felt that their organisations were understaffed, which would complicate their participation in a research project. Gaining access to the field therefore demanded much effort, and the project members had to take care not to ask for additional time and effort from the CSOs. Once the researchers were immersed in the fieldwork, and trust was built, we heard and experienced less reluctance as we continued to reflect on how to ‘give back to the CSO’, which in some cases resulted in our giving token gifts to interview participants, such as vouchers, or paying for a drink or a meal or buying chocolates.

As this project consists of a nested structure – namely recruiting participants that take part in activities organised by a specific CSO – additional care was needed to ensure safety, privacy and the application of ethical guidelines. Some CSOs work on very specific topics (such as gender-based violence or poverty) with a particular target group, which may be newcomers from a specific region, people confronted with violence, or refugees. To reflect on how we would deal with ethical considerations, especially to guarantee and safeguard anonymity and pseudonymisation of the data when desired, we used a ‘decision tree’. This decision tree was followed by researchers to understand the level of anonymisation and pseudonymisation needed to be used to safeguard anonymity – even if names of individuals and organisations are not disclosed. This was needed due to the selection of the CSOs in our sample which may be easily recognisable by virtue of their work in a particular context and country. In addition, given that many of them were relatively small, functions of participants (for example, CSO director) or socio-demographic information (for example, an older Pakistani woman working in a bakery) could immediately identify participants. Hence, further measures were taken to not disclose the personal and sensitive information of participants and CSOs involved. Another precaution taken was to ensure that if participants wish to disclose information about themselves, this would not impact the anonymity and pseudonymisation of other participants. The latter is especially important concerning the photo-trajectories and to avoid a paternalistic approach to anonymisation (Kara, 2022).

In Fig. 5.1, we give an overview of the considerations that the researchers, and the consortium, followed to decide on how CSOs and their participants will be discussed and referred to in our research outputs.

This figure shows that the following elements are considered: (1) the size of the CSO, (2) the specificity of the data, (3) being able to track down the quote to one person and (4) the position of the participant in society.

1. Is the CSO large in size? The size matters to ensure people cannot be identified based on their demographic data when revealing the name of the organisation. This could mean, for instance, that approximately 10 persons with similar demographic data are needed to not automatically disclose information when participants do not wish to. If not, the quote used for a participant may be anonymous on its own but not when the name of the CSO is mentioned.

2. Is the quote itself very specific? If the information revealed from the quote is very specific, you may immediately guess the organisation this person is part of.

3. Is the quote clearly tracked down to one person? Discuss this with the participants first (and thus discuss their anonymity) or write it in such a way it is not formulated in a quote but part of a larger chunk of text.

4. Does the participant hold a vulnerable position in society? Discuss the differences between anonymity and non-anonymity with the participant. Discuss the risks they could face and support them when making decisions on participation and/or anonymity.

Regarding anonymisation of CSOs, we have not, so far, encountered any major challenges. However, we have reflected on the need to consider this when talking about the ethnographic process in conferences or even to our advisory boards. Given the diversity of CSO landscapes in all the countries involved in our consortium, it is also relatively easy to identify organisations based on their characteristics, scope and target population. Particularly in relation to smaller countries like Austria and Belgium, we have decided to talk about CSOs in more general terms, always being aware to not give away any information which could identify these organisations. Anonymisation of the CSOs was considered essential both for the analysis and presentation of findings about the opinions and discourses of their staff members, as well as for better anonymisation of participants from the migrantised groups with whom these organisations work. As for women from the target groups who took part in in-depth photo-elicited interviews, we found that it was essential for us to spend some time together with them, even before doing the interviews, to explain the aim of the research better and gain their trust. In other words, anonymising organisations was deemed to be necessary for photos to be shared – especially when these touched on gender empowerment, which is a sensitive and highly debated topic in society.

Our negotiations with participants – often combined with their participation during the interviews – usually followed a stepwise process. Already in our consent form, participants had the option to agree to either one or several aspects of the project. This ICF included an option to stay anonymous, to use photos during the interview and to use the photos for the art exhibition and/or photo-essay book. Given that this was already quite detailed, somewhat academic in nature and formulated in the dominant language of the country, presenting and explaining the ethical procedures took quite some time. As participants were not fully aware of what would be entailed at the outset, a stepwise approach to participation in the research project was agreed, with participants first being asked which elements they wanted to participate in: this felt reassuring and helped to overcome initial barriers to participation. It often helped that the researcher took part in some activities organised by the CSO before conducting the interview, building rapport with the participants and gaining mutual trust. Additionally, we found that having WhatsApp groups for researchers, participants and CSO staff was another medium to build rapport and provided a way for participants to get in touch using direct messages to raise questions concerning the use of their photos or objects and potential references to their name. Within these groups, only members of the same CSO were included, and no sensitive data were shared. Rather, practical questions about the research were shared. We also noticed that researchers with a migration background experienced less difficulty with reaching participants. One Belgian researcher with a migration background also mentioned that speaking the same mother tongue or having the same ethnic or religious background as the participants made it easy for them to be open and make connections. After getting to know the researcher and showing their first photos during the interviews, it seemed easier to also agree to contribute to the art exhibition at a later stage. During the first interview, the researcher explained the assignment and which kind of photos or objects the participants were expected to bring during the next interview.

Using an interpreter, as we did in Austria, offered another approach to establishing rapport and facilitating a clear understanding of the consent forms. Interpreters were also from migrant backgrounds and shared both language and cultural background with the participants. They explained both the process and the different steps of the interviews, as well as the implications of agreeing to participate with quotes, pictures and sharing the pictures publicly. This was treated as a separate stage, before the interview, and was conducted directly in the language of the participant, explaining and replying to questions, without translating them back to the researcher. This helped to create a sense of intimacy and trust. Some of the participants shared after the interview that they felt comfortable to ask further questions about the use of photos, for example, at a later stage during the interview, without feeling embarrassed that they did not understand the idea for the exhibition during the first explanation, because they already trusted the interpreter.

Using an interpreter, however, does raise additional questions pertaining to ethics and power relations and to levels of sharing and opening up during an interview. As also noted in other analyses (Bowlby & Day, 2018), using interpreters during interviews can change the dynamics and sometimes even partly the content of the interview. We found that interpreters with a similar background to that of the participants built a different type of rapport with the participants than researchers who do not speak the same language. The use of interpreters can also create ease and more openness in the interview, since it avoids the sometimes intense dynamics of a one-to-one interview. Overall, we have found that having an additional person present during the interview with shared cultural and linguistic knowledge with the interviewees created a different setting and changed the power relations and positions in that context. A shared cultural background and language is, of course, just one of the possible common denominators. In our case, gender between the participants and the researchers was also the same during all interviews, having only female interpreters. But other socio-demographic factors, such as social class, educational background, rural/urban location and age are all potential additional points of difference that varied and impacted the flow of the interview. Another contentious point about having interpreters involved relates to the way in which interviews are translated, whether immediately (literally) or after a duration of time: these dynamics can affect the flow of an interview or in some cases leave the interviewer in a position of partial understanding. Although we did not translate them afterwards for further triangulation, this would have further added to insights into the concepts used. While in some interviews, this could have facilitated the interview, it might, in others, have disrupted the flow. Despite these difficulties, this does not necessarily undermine the quality of data obtained (Bowlby & Day, 2018).

Given that our research project takes a very personal and even intimate look into each participant’s understanding of womanhood and empowerment, our approach to the interviews was to give participants a large role in deciding which research topics they wanted to delve into. In setting-up the interviews, we offered them space to reflect and discuss freely their selected topics, to go in directions that they felt are important and to share life details only to the extent that they felt were relevant for the topic. Since this is at the core of the participatory approach we aim to take, we found that the changed dynamics brought about by the interpreter contributed to a direction of participants feeling comfortable and well understood, even if the flow between questions and replies might have been influenced. To ensure the reliability of the information shared, we tried out several methods for presenting a shorter and a longer translated version of the ICF, a short description of the project and of the interview process, that can be taken home by the participants. We also analysed the translations of the interviews afterwards. In addition, we used the same interpreters for each language, to ensure consistency.

Depending on the background of the participants, the time taken to explain the project and sign the consent forms varied. Completion of the consent forms took less time for the CSO organisers, compared to the other participants. For some of the women with lower education, or lower levels of literacy, it was overwhelming to read this document. What came across as difficult in some cases was that while, on the one hand, we explained that we will keep their anonymity and privacy and we will not disclose their names, unless they explicitly agree, on the other hand, we asked them to identify themselves when signing the consent form. We spent time carefully explaining the way we stored the consent forms, and that they would not be connected to the analysis in any way.

We explained that signing is a way of guaranteeing that we would share only the information that they agree upon, which kept the researcher responsible. For instance, in Vienna, one woman asked whether she could see the answers of other women, and we explained that this was not possible, precisely because of privacy issues, which, in fact, convinced her that we would not disclose sensitive information. This, we found, was particularly important for women who were going through divorce procedures and feeling vulnerable. Overall, it felt that taking the time to explain and reassure participants that their answers would remain anonymous and to explain the process of analysis was beneficial for trust building. This was particularly important in the cases where women shared traumatic life stories.

In the case of our work in Belgium, participants experienced it as discouraging if we presented the consent form right away – as this might seem to be putting too much emphasis on ‘concerns’ rather than on what this could also mean in terms of ‘safety’ and ‘opportunity’ for the women. For this reason, the researchers in Belgium sometimes presented the consent form at the end of the first interview after providing general information on ethics. In Austria, the researchers found that taking the time to explain the project, to describe the interview process and to explain how anonymity and privacy would be secured was a step towards establishing trust. While this did take significant time at the beginning of the first interview, it demonstrated to the participants the two-way engagement and the willingness of the researchers to invest their time to reassure and make them feel comfortable. This discussion was used as a time to discuss possible concerns and worries about sharing personal stories and details. Thinking of the informed consent procedure as a significant step in the interview process, rather than as a formality, allowed researchers and participants both to understand the details and to better understand the meaning of research.

Questions about ownership and power relations in visual research methods emerge even before the research project starts. While claiming that the usage of visual methods can result in more empowerment, awareness and ownership by so-called ‘hard to reach’ groups or vulnerable groups in society, many research projects design their research questions without consulting their target group. This brings in some power imbalances, as the research questions do not always resonate with the questions, demands and voices of the people involved (Milne & Muir, 2019). Within our project, CSOs were involved in the proposal writing and helped to recruit participants who in some cases already had a desire to discuss gender and empowerment in society. When the broader research design was set up, the interpretations of the civil society partners and their participants were warmly welcomed, albeit that this could have resulted in some selection bias of participants. Hence, solely those attracted to the theme, or the specific CSO focusing on this topic, engaged with this time-intensive project and met several times with a researcher to further develop the interview questions.

Concerning the usage of visual methods, many debates arose regarding how we should prepare participants when taking and discussing photos on gender empowerment. These debates concerned professional support for training when using visual methods or even hiring a professional photographer to re-enact the photos with the participant. Arguments that were in favour of professionalisation stated that everyone deserves to showcase photos of high quality in an art exhibition, and this would also be a benefit from participating in this research project. Concerns were related to the impact that the art exhibition might have on the participants if we did not manage to come out with photos which, when exhibited, made them feel proud of their process and work, making them feel exposed or vulnerable. For many participants, the process was experienced as rewarding, so instead of focusing on the end result, many focused on this participatory experience as being reflective and meaningful. After one of the final interviews, a participant informally discussed with the researcher that she wanted to edit or change some details about her picture because ‘after knowing the project better and understanding the bigger purpose of representation’ she learned that she wanted to ‘pass on the message to society of different generations of migrant women and that all have made a difference’ (ORG3_BE_10). This conversation continued with the participant stating that, for her, the message or story behind the photo is more important. Arguments against professionalising the process of taking photos underlined that the participants should be valued in their own right, and that seeing the photos they had taken themselves in an essay book, art exhibition or when discussing them with a researcher was empowering in itself. In the end, both strategies were experienced as ‘empowering’. In addition, many participants wanted to show their names when presenting their photo in the art exhibition, although this did cause additional efforts for the researchers in terms of anonymising both the organisations and participants’ quotes in academic publications. Finally, during the selection of photos for the photo-contest, we differentiated between those who enjoyed training, workshops or had a professional background in photography.

The procedure and impact of the photo-eliciting trajectories also prompted some ethical reflection. Participants showed photos of themselves during turning points in their lives – working, obtaining a degree, marriage, first child, migration – and reflecting important aspects of their identities and cultures, such as a book or an inspiring individual, and life philosophies. In some cases, photos felt very personal and were therefore harder to bring to the interview moment. For this reason, we opened up an option for participants to also bring objects, as this may feel safer and so make the selection easier. Some brought (a photo of) an old but special newspaper or a memory box. This corresponds to the remarks made by Luttrell (2010) and Milne and Muir (2019), who question the notion that all research participants welcome the use of camera technologies to express their stories and ideas.

While we emphasise the participatory nature of our project, we must acknowledge that only one ‘side’ is asked to provide photos (both as a research tool and as an output). One could question this approach, arguing that it does not provide an equal context to participate together in one research project (Van Praag, 2021). Following the conventions of interviewing, the researchers did not provide photos, nor did they discuss their views and ideas on ‘empowerment’. Even though we tried to personally engage with the photos and the corresponding narratives, we are aware that less attention had been paid to reciprocity during our interviews. In some cases, we tried to balance this by discussing the challenges and the different ways of approaching the task, with the researcher and/or the interpreter sharing what they would bring to such an interview and what it would mean to them, with the aim of creating more of a shared space of exchange and reciprocity.

Second, given the contextual difference between the photo-eliciting interviews and the art exhibition, we encountered some difficulties when participants used photos of themselves and/or their family members. As using the photos or objects during the interviews – so as to show how proud they are of their family – had less ethical considerations than exhibiting these photos or objects for the public, this further complicated reflections on the limits to anonymity in this context. Added to this, minors were often included in the photos, which we only displayed in the exhibition if explicit consent had been given by one of the parents for their display. As for others shown in the pictures, direct consent was also obtained from these people.

Third, trauma often came up for the researchers during the photo-eliciting interviews for various reasons. For instance, migrantised researchers felt that some topics made them emotional, for example, when discussing first generation parents and their similar struggles. Both in Belgium and in Austria, it was noted that participants appreciated it if the researchers expressed and shared some of their experiences. In this respect, the sharing of some personal details by the researcher or by interpreters provided a sense of reciprocity and better understanding. On some occasions, the researcher experienced emotional challenges to process the often painful stories told by the participants.

Importantly, and related to the third point, many participants also experienced emotions and trauma when selecting and sharing their photos. For instance, one participant mentioned liking these kinds of challenges to their creativity, which helped her to reflect on what this concept really meant to her, to visualise it and to discuss some of her emotions:

Showing photos and simultaneously having the opportunity to engage in a conversation around that photo was well received by the participants. As other interviewees put it:

The choice of photos often resulted in some stress to pick the ‘correct’ photo or to put into photos what a vague concept such as ‘gender empowerment’ means. One person said jokingly that she encountered difficulties picking one photo: ‘Oops, which photo, this one is good or that one’, adding afterwards that she didn’t find it easy to express herself using photos ‘maybe I can present myself better’ (CSO1_BE_7). Yet, she said that hadn’t had much time to prepare, which affected her choices and the experience in general. On a similar theme, other participants said that ‘it was interesting but at the same time also difficult as I couldn’t find what I mean myself’ (CSO2_BE_6) and commented on the wider meanings of such difficulties:

While the experiences with photo-eliciting interviews varied between participants, we can say that in many cases, these interviews did bring up some valuable reflections, both for the participants and for the researcher.

Some participants wanted to show even more and others deliberately searched for photos that represented specific moments in their lives. One found that ‘for me, it’s easier with photos, as you will focus more on searching for specific moments in time. And at the same time, memories come up, that goes easier with photos for me’ (CSO1_BE_5). As reflected in the following quote, this shows the complexity of affect and emotion, as both nostalgic feelings and also more traumatic experiences came up.

Participants wanted to make sure that several photos were included as part of their narrative of empowerment. A ‘before and after’ was indicated by selecting a photo illustrating the way the participant felt before and after a significant event, such as separation, which illustrated a path towards empowerment. This participant insisted her photos only made sense together and must be used as two parts of one illustration. Taking into account such a requirement was also considered part of our ethical responsibilities.

Finally, some comments were made about ‘selecting photos that can be shown to someone else’. This suggests that some participants considered representations of themselves, or the group they belong to, when selecting the photos, as described in the following conversation with a participant (CSO1_BE_3):

This extract reminds us that there is a huge variety of ways that photos can provoke emotions and stimulate the conversation, as well as sometimes limiting the sharing between the participant and the researcher.

This chapter sheds light on the use of visual methods when conducting research with migrantised women in five European Union (EU) countries as part of our comparative project. Visual methods are often used to reach out to specific groups and discuss sensitive topics or topics that demand fresh and novel insights (Duffy, 2011; Epstein et al., 2006; Glaw et al., 2017; Gube, 2022; Wang & Burris, 1997). In our case, we aim to deconstruct the concept of ‘gender empowerment’ by using a so-called empowering technique, in which participants are free to show a novel perspective on this concept, with time to reflect on the topic. This concept is frequently used in research – and recurs in many policy initiatives – and yet is hard to translate or ‘capture’. Using photo-eliciting techniques could therefore provide an elegant way to further define and conceptualise this concept, by exploring visual materials with participants.

In the course of our project, we realised that at different stages, ethical considerations came to the fore, which needed careful thought and often did not have clear-cut solutions. We realised that these ethical aspects needed to be negotiated before, during and after the photo-eliciting trajectories with interviewees. Furthermore, debates about who owns the research, and about the photos and the people in them, were held, reflecting on the need to include a sense of ownership into this research trajectory as well as the role professionalism plays in this context. These debates reflect the critical insights offered by Milne and Muir (2019), which highlight that giving voice to so-called vulnerable or ‘hard to reach’ groups in societies may challenge existing ideas about ownership and how social action takes place.

Finally, the use of visual methods, accompanied by the time that we invested in this project, was associated with participants and researchers becoming emotionally engaged with the project. This raised concerns about peoples’ well-being, which continue throughout the research and will remain even after it ends. We hope that by shedding light on the experience gained by our project, we can further inspire future researchers to (better) reflect on the ethics when designing and executing research using visual methods.

Some future research recommendations can be formulated, especially for university ethics boards. We recommend that research institutions should, where possible, have similar ethical forms and requirements for ethics review and should formally acknowledge the decisions made by ethics committees from other institutions. Second, this collaboration should facilitate the development and sharing of a set of ICFs in simplified language and in different languages across research institutions, to avoid overburdening participants with unnecessarily complex material (Crow et al., 2006). Finally, ethics boards and advisors should implement a processual approach to ethics in several moments in time of a research project and develop clear reflection guidelines. Although the literature on ethics may already point to these solutions, they are not widely implemented in research institutions, albeit that the demands and ideas of ethics boards vary considerably.

We would like to thank the reviewer for the thorough feedback on this chapter. We are very grateful for the language edits made by Helen Busby, which significantly improved the quality of this chapter. Lore Van Praag would like to thank Elien Colman for her reflective insights on the use of art for research and mental well-being purposes and on how photographs can say more than a thousand words. This chapter is based on the findings of the Horizon Europe ReIncluGen project, which has received funding from the European Union’s Horizon Europe Research and Innovative Action under Grant Agreement No. 101093987.