WhatsApp as a Tool for Researching the Everyday Lives of Venezuelan Refugees Settling in Brazil

In this article, we explore the role that WhatsApp can play as a research tool for investigating the experiences of settling refugees. Messaging apps can help researchers collect data about people’s everyday lives while also providing insights into processes that are difficult to study as they happen. The communicative affordances of messaging apps also facili‐ tate spontaneous interactions in research and the flexibility needed when working with mobile groups of people, such as refugees. We build on our experience of interacting together with Venezuelan refugees through the Conexión Sin Fronteras (Connection Without Borders) WhatsApp group, which was designed by the researchers in the form of an intervention focused on building community among Venezuelans settling in the city of Boa Vista, Brazil. Our experience shows that data collection in WhatsApp allows researchers to obtain relevant insights into social support, relationship‐building, and negotiations of rules in a group context. However, our research outlines challenges related to the varied engagement of participants in WhatsApp group chats and the difficulty for researchers to be always present during group conversations. Limitations to the use of WhatsApp in research with refugee populations also include restrictions in terms of internet capac‐ ity shaping the types of data participants choose to share. In this context, it is crucial to address barriers to access to con‐ nectivity and create opportunities to enhance refugees’ literacy regarding data collection in digital spaces. We hope these findings will contribute to the development of inclusive methodological approaches using mobile apps in refugee settings.


Introduction
The proliferation of smartphone devices and access to affordable mobile connectivity have facilitated the adoption and use of mobile messengers among people globally. The multimedia affordances of smartphones and messaging apps allow users to send text, share real-time locations, voice recordings, pictures, emojis, GIFs, hyperlinks, documents, and videos. Due to their deep integration into everyday routines, messaging apps can be seen as useful tools to research people's lives, experiences, and interactions in a variety of contexts . A growing number of studies have identified that mobile instant messaging apps can widen participation of population groups that are hard to reach or involve in research given their vulnerable social and economic situation (Kaufmann, 2018;Marzi, 2021).
The present article explores the adoption of WhatsApp when working with refugee communities. Specifically, the analysis builds on our experience of interacting together with Venezuelan refugees through the Conexión Sin Fronteras (Connection Without Borders) WhatsApp group, an intervention project aimed at fostering social community building among the participants. Interventions can be defined as "purposively implemented change strategies" developed to accomplish a desired behavior or outcome at the individual, group, or community levels (Fraser & Galinsky, 2010, p. 459). For the intervention in this study, we adopted a longitudinal design with WhatsApp as a research tool for accompanying the lives of Venezuelan refugees in Brazil in order to analyze how their engagement in the WhatsApp group can support their settlement process. In this article, our focus is not on evaluating the intervention or presenting the changes in participants' life trajectories during the study, but on the possibilities and limitations of using WhatsApp for studying the refugee experience of settlement. This article begins by providing a brief overview of the research context and studies on the use of mobile apps for data collection. It will then go on to develop a more detailed account of the project, followed by the analysis and conclusions.

Contextualizing Venezuelan (Forced) Migration in Brazil
Since 2015, four million Venezuelans have fled hunger, violence, and hyperinflation in their country. Official statistics estimate that, by the end of 2021, 305,000 migrants from Venezuela arrived in Brazil (International Organization for Migration, 2021), with the majority crossing the border city of Pacaraima and proceeding to Boa Vista, capital of the state of Roraima in northwestern Brazil. The legal situation of Venezuelans is determined by two "regularization" routes of the Brazilian government, which allow them to apply for a residency permit or refugee status (International Organization for Migration, 2021). Despite awareness of the complexity and diversity of circumstances shaping both legal pathways for Venezuelans, this article utilizes the term "refugee(s)" in the broader context of forced displacement in an effort to include as many perspectives and experiences as possible. Brazil is regarded as a country with pragmatic, humane, and progressive asylum and refugee policies, which include entitlement to work, freedom of movement, and access to social services (Martuscelli, 2021). In practice, however, Venezuelans do not have access to a state program that supports their integration trajectory, and many depend on the large support network of international, national, and local organizations and networks to alleviate their precarious conditions of life in Roraima. These challenges are reinforced by Venezuelans' inability to sustain continuous and reliable digital access and connectivity that can facilitate their access to information, job and educational opportunities, health assistance, social connections, etc. (see Alencar, 2020). Within this context, it is important to highlight that one in two Venezuelans living in Boa Vista does not feel sufficiently informed and 69% do not have Wi-Fi access (R4V, 2020).
Digital access and use among refugees are conditioned upon various contextual factors. The lack of support from government and humanitarian organizations regarding the provision of adequate digital infrastructures (Witteborn, 2021), as well as restrictive telecommunication policies can impact refugees' access to digital connectivity and spaces (United Nations High Commissioner for Refugees, 2021). On the other hand, the needs and experiences generated from the migratory process have led refugees to create tactical forms of technology use. Recent evidence suggests that Venezuelan refugees share their cell phones with family, friends, or acquaintances, motivated by their need to communicate with family members, build social relationships, search for administrative information, and in some cases, transfer money to Venezuela. This study considers the precariousness and social and digital inequalities affecting the lives of Venezuelans in Boa Vista, factors that require ethical attention and care from researchers when it comes to the implementation of methodological approaches using mobile apps in research with marginalized communities in diverse settings.

Using Mobile Messaging Apps as a Tool for Data Collection
Messaging apps have a demonstrated capability to address the methodological challenges of researching people's everyday life contexts and "real-time" experiences (Kaufmann et al., 2021). Mobile messengers enable people to establish individual and collective connections at any point in time and across different time zones (Mols & Pridmore, 2021), with the possibility of collecting audio-visual and textual data that can be shared online between participants and researchers and within research teams (Jailobaev et al., 2021;Marzi, 2021). The new opportunities of using mobile messaging apps for qualitative in situ data "can provide insights into processes that are otherwise difficult to study, or that can be lost if accessed in retrospective accounts only" (Kaufmann & Peil, 2020, p. 231).
There is a growing body of literature that recognizes the affordances of smartphone devices and messaging apps for data collection and participant interaction in qualitative research Manji et al., 2021). A significant analysis and discussion on the subject were presented by Kaufmann et al. (2021;Kaufmann & Peil, 2020). In their study, the authors accompanied young adults throughout their everyday life contexts and interviewed them about their daily routines and momentary experiences, while prompting participants to employ the various multimodal features afforded by WhatsApp (e.g., sending texts, voice messages, pictures, videos, emojis, GIFs, etc.) to express themselves. In contrast to quantitative methods for mobile apps that adopt standardized questionnaires in fixed schedules, Kaufmann et al. (2021) highlight the potential of mobile messengers to offer spontaneous interactions between participants and researchers, making it a flexible and open research process that is characteristic of qualitative research (p. 3). Considering the ease and routinized use of mobile phone apps combined with their near-synchronicity, these technologies can allow interactions to take place in a less intrusive way (Jailobaev et al., 2021).
Furthermore, spontaneous interactions using mobile phones can reduce power imbalances and inequalities in research (Nash & Moore, 2018). In this regard, Dawson et al. (2020, p. 212) highlights that instant messaging apps should be used as a method to promote "friendly and natural-feeling conversations," as this approach helps balance the participant-researcher power relations, enhancing the voices of the groups under study. Previous research has stated that flexibility of data collection via messaging apps, such as WhatsApp, facilitates the recruitment of participants regardless of their geographical location, while also retaining contact with them should they move to other places Kaufmann & Peil, 2020). This is especially true when considering the participation of mobile groups of people in research, such as refugees (Kaufmann, 2018). Researching refugees' everyday lives can be challenging due to the volatile conditions of their migratory trajectories and situations (Beduschi, 2018). Several factors related to the lack of stability in the host communities, temporality of living conditions (e.g., living in temporary camps), as well as the difficulties that researchers experience to access social networks of refugee migrants make them one of the most hard-to-reach population groups (Shaghaghi et al., 2011).
To date, only a few studies were found that have adopted mobile messaging apps in the context of datagathering strategies among refugees. With an emphasis on the analysis of refugees' daily practices of digital media use, messaging apps have been used by researchers as digital tools to stay in contact with refugee participants and to conduct interviews (Kaufmann, 2018;Marlowe, 2019;Twigt, 2018;Zijlstra & Van Liempt, 2017), as well as part of the ongoing participatory research (Godin & Donà, 2021;Leurs, 2017). A recent study by Palmberger (2022) involved the use of digital diaries among refugees to record their digitally mediated care practices through screenshots from WhatsApp communication with family members. Despite the importance of these approaches, there remains a paucity of evidence on the methodological potential of messaging apps for researching refugee processes. This article is contributing to filling this gap in knowledge by examining the longitudinal application of the mobile messenger WhatsApp within one intervention study with Venezuelan refugees in Brazil.

Study Design: The Conexión Sin Fronteras WhatsApp Group
The WhatsApp group Conexión Sin Fronteras (Connection Without Borders) was designed by the researchers in the form of an intervention focused on building community among Venezuelan refugees settling in the city of Boa Vista. This intervention aimed at analyzing how participants' interactions in the WhatsApp group could potentially enhance their experience of settlement. Our decision to use WhatsApp as a research tool in this project was motivated by the prevalence of the mobile app as a means of communication within the Venezuelan community in Brazil. Recent reports revealed that 39% of Venezuelans have indicated their preference for WhatsApp when it comes to receiving information (R4V, 2020). The possibility of utilizing the attention participants routinely devote to their most relevant digital space enables researchers to accompany them through specific life experiences (Kaufmann, 2018;Kaufmann et al., 2021).
We recruited 18 Venezuelans with different housing and living conditions in Boa Vista. They lived in rented apartments, refugee shelters, settlement occupations, and in some cases, were in street situations. Most of the participants were unemployed, while others were doing small daily services in the building and construction sectors, informal jobs, care, or domestic services. None of the participants had a mobile phone or a smartphone when they arrived in Brazil. Some reported having to sell their phones in order to pay for their trip to Brazil, whereas others left the devices with their family members so that they could communicate with each other. At the time of the research, all the Venezuelan participants attended Portuguese classes and a computer lab program at the Technological Reference Center of the city's public university in the context of local integration initiatives for migrants and refugees. As coordinator of the technology center, the second author of this article has been actively involved in a range of initiatives aiming at improving refugees' access to ICTs, which helped in our engagement with the community. We started recruiting participants in the first week of lessons at the center, as it was important for the project's approach to community-building that they did not know each other yet. Through short informative sessions during class time, researchers were given the opportunity to present the study to potential participants, explain their role as researchers, and invite them to participate in the project. The recruitment process resulted in a diverse group of participants (e.g., in terms of age, gender, educational levels, disabilities, sexuality, etc.), making it possible for the inclusion of multiple perspectives and narratives in the analysis. The sample included 10 women and eight men, aged from 18 to 64 years old, and living in Brazil between two months and two years. There were also differences regarding levels of school education (from incomplete first years of education until graduation) and categories of marital status (single, married, widowed) among participants. For this study, pseudonyms are used as names to protect the privacy of participants. We also employ the expressions "group members" and "group participants" to highlight the way Venezuelans referred to themselves throughout their engagement in the WhatsApp group.
As part of the intervention design, we organized introductory sessions to provide participants more details about the project, answer their questions and give them sheets and informed consent forms in Spanish, approved by the Ethics Review Board of the Federal University of Roraima. These sessions were scheduled at two different times to best accommodate the availability of participants. Considering power imbalances that occur in the design of projects with refugees and our own position as privileged researchers in both Western and non-Western academic settings, we aimed at engaging in a relationship of reciprocity with participants through which both groups would be responsive to each other's needs during and beyond the research process. In the context of postcolonial indigenous methodological approaches, the principle of reciprocity implies that "the research must be mutually beneficial for both researchers and the local communities while listening to participants' voices and considering their needs and goals" (Restoule, 2008, p. 203). Based on this perspective, we claim that the relationship between the researchers and community members in this project was transactional, meaning that we provided each participant with a smartphone and basic data packages to use the internet every month and in exchange the participants provided the researchers with access to the personal information shared via the WhatsApp group. The smartphones used in this project were donated by the UN Refugee Agency in Boa Vista and the costs of the internet credits (€3,40 per month to each participant) were covered by Voorbereidend Jaar Erasmus, an educational program at Erasmus University Rotterdam supporting sociocultural and digital inclusion of refugees in the Netherlands. The provision of the phones and funding for the material costs was made possible through the researchers' collaborations with these organizations in previous initiatives. By the time the project was completed, the researchers informed participants that they could keep the smartphones for themselves.
The project started officially in January 2019 and spanned eight months. We first conducted interviews with the participants to familiarize ourselves with their stories, communication practices, and everyday life. After participants' smartphones were set up, we created the Conexión Sin Fronteras (Connection Without Borders) WhatsApp. It was envisioned that the WhatsApp group would serve as a collaborative space where participants could send messages to the group at any time and share experiences related to their everyday lives in Brazil. This accords with intervention approaches to WhatsApp groups in health research that used the technique of free (unstructured) and participant group observation (researchers were also members of the WhatsApp group), allowing researchers to immerse themselves in the research process and learn directly from participants' experiences in the groups (Arroz et al., 2019;Henry et al., 2016). Arroz et al. (2019) stated that the fact that researchers are members of the WhatsApp group can potentially lead to bias in the research process and provided ways to minimize this risk, such as active participation of researchers in clarifying questions raised by participants as well as in raising questions about participants' interactions in the group.
In this study, we, the researchers, positioned ourselves as members and facilitators of the group, stimulating unstructured conversations and posts while proposing questions associated with participants' experiences and situations throughout the project. Every week, we asked participants questions about their settlement processes that involved the use of their mobile devices. Additionally, we asked them to further elaborate on the main issues, situations, and information that they shared in the group. The topics ranged from information on basic needs and support, jobs, romance, ways of coping, and political discussions, to name but a few. Participants were encouraged to use the whole range of WhatsApp features (send pictures, videos, emojis, screenshots, links, etc.) for their responses (Kaufmann, 2018;Kaufmann et al., 2021). There were times when we needed to remind the group to answer the questions of the project while retaining flexibility due to the challenging circumstances of participants' lives. In total, data from the WhatsApp group generated 341.4 MB in storage, with 921 media posts (image, video, and audio files), links, and documents.
Access to the group was strictly limited to the participants and researchers in the project. Since the inception of the WhatsApp group, all participants were properly informed by the researchers that their information posts and interactions on WhatsApp would be used for academic purposes. Standard ethical procedures (confidentiality, consent, data protection, etc.) were complemented by a practice-based ethics approach that allowed for constant negotiations of interactions among group members (Møller & Robards, 2019). In this way, ethical considerations became part of the entire WhatsApp group project and not just a side activity (Marino, 2020). Through an open and collaborative space, we decided together that the contents of posts and interactions from participants would not be shared outside the group.
Within the study, we also organized monthly meetings in person at the Technology Center to discuss further participants' experiences and specific topics that concerned the entire group. The integration of WhatsApp in a qualitative multi-method research design allowed for enhanced insights into the collaborative and learning process of participants as a group (Arroz et al., 2019;Henry et al., 2016), following the exploratory and participatory objectives of this project. In the last month of the project, we conducted interviews with participants about their experiences of engaging in the WhatsApp group and how this shaped their settlement processes. Given that the first author was based in The Netherlands and the second author in Brazil, the organization of the project among the researchers has been mostly managed via WhatsApp, except for the initial and final interviews, which were conducted by both researchers in situ. Drawing upon the findings emerging from this study as well as empirical discussions of messaging apps as research tools, the next sections analyze the strengths and weaknesses of using WhatsApp as part of a longitudinal method for gathering data and studying the lived experiences of members of the Venezuelan refugee community while settling in Brazil.

Strengths of Utilizing WhatsApp for Data Gathering When Working With Refugee Settlement Communities
The WhatsApp group generated a sense of community and feelings of belonging among participants. The Venezuelans in this study appreciated the possibility of being connected to new people as well as learning about their professional and personal developments throughout the research process. As one participant said: "The project was very good because many of us met new people, had opportunities and doors opened to many friends. I saw in the group that many of my peers are working, taking courses, doing other things" (Pilar, a 44-year-old woman, doctor, waiting for the regularization of her diploma to be able to work). For Laura, (a 29-year-old woman, mother of three, housewife), the WhatsApp group project brought opportunities for the participants as well as for researchers beyond the research project: It helped me a lot as a person. It was very innovative to participate in something like that [WhatsApp group research] because I never saw this in Venezuela and here I had this opportunity….Because I know that it will be of some use to me or to you [referring to researchers] as well.
From the outset, participants were engaged in jointly creating a community by starting their own conversations. This offered the researchers the possibility to observe how conversations unfolded in the group, which might have been different if the researchers had initiated the conversations. Interactions in the group started with participants introducing themselves and saying how they first used their phones. The group's first post was from Josiane, a 19-year-old woman, pregnant with her first child, who shared in the group a screenshot of the message she sent to her mother saying: "Hi mom, I'm already connected to you" (Figure 1). Josiane has not spoken to her mother since she arrived in Brazil, two months earlier. Other participants interacted with Josiane's message and shared their experiences about their first phone use and feelings of missing their family. This group interaction occurred in a spontaneous way, allowing everyone to participate as they would like, including, us, the researchers.
Considering the engagement of participants in the WhatsApp group for the benefit of community building, many were willing to share with the group information about their everyday life, which they knew was also being shared with researchers. In one case, Mr. Edward (a 63-year-old former teacher from Bolivar state) shared humorous material showing his daily routine taking care of his granddaughter in one of the refugee shelters. As shown in Figure 2, members reacted by making jokes, which provided the researchers with an opportunity to access momentary but genuine feelings of happiness experienced by participants.
At the same time, the development of spontaneous interactions among group members enabled by the informal, free-and-ease communication style (e.g., light responses, emojis, visual illustrations) also allowed researchers to obtain insights into moments of anxiety affecting participants' experiences (Jailobaev et al., 2021). For example, a WhatsApp snapshot in Figure 3 shows that group members sent messages of encouragement, support, and tranquility, especially with the use of emojis, when one of the members said he was feeling nervous and anxious prior to his job interview.
The WhatsApp group became a relevant space that participants used to obtain and share all kinds of information. The group was often populated by messages  and website links related to job opportunities, education, health, legal procedures, recipes, cultural celebrations, and news from Venezuela, functioning as a source that participants had access to daily. Within this context, it was also possible to observe that the group space was used as a mechanism for collective checking of contents that raised suspicion among participants. In one of the cases, the group verified and judged as false the information about food donations by a supermarket chain in the city, and in another case, the group also analyzed the news on the plan of invasion of Venezuela by American troops to remove the Venezuelan President Nicolas Maduro from power. In this study, the design of the WhatsApp group in the form of an intervention focused Message 1 (from a group member): Hi everyone, there are 10 people leŌ for my turn in the interview and I feel [despair and nervousness face]. Asking God that everything goes well.
Message 2 (from a group member): Brother, rest assured, have faith in God and in yourself, what is for you, it's for you, relax, this job vacancy is yours.

Message 3 (from a group member):
Whis you the best of luck, don't worry, everything will be alright. Let´s go! (ParƟcipant name) [contentment face]. on building community among resettling refugees facilitated relationship building and support that extended beyond the app. Some Venezuelan participants reported helping their peers by accompanying them to places, such as hospitals, banks, government offices, or training for job interviews, while also taking photos of themselves during these moments through their mobile devices and sharing them in the group. Aside from bringing direct benefits to the participants, the group also enabled us to better understand participants' immediate requests for help and provision of assistance in ways that would not have been possible had the researchers adopted another method of digital ethnography (e.g., conducting individual WhatsApp chats with participants, or following participant interaction on Facebook groups).
During our analysis and reflections on the use of the WhatsApp group in this project, we came to realize that conflicts in the context of the WhatsApp group can function as a means for further observation of how relationships among refugees resettling take place and are being negotiated in real-time. Management of participants' expectations regarding the care and attention of other members emerged as an important issue. Compared to group relations in other networks such as Facebook, WhatsApp group dynamics can generate more intimacy and dependencies, shaping members' expectations about responsibility and pressure to respond (Mols & Pridmore, 2021, p. 3). One of the most emblematic situations experienced in the group involved the temporary departure of one of the participants who reported not receiving sufficient attention from others after going through a delicate health problem. Yannine (28-year-old woman who arrived in Brazil with her boyfriend and his family) posted several messages in the group about the alleged lack of support, calling members insensitive and saying that she thought the group was like a family to her, but there was actually a relationship of hypocrisy among group members, except for just a few. Yannine's attitude towards the group led to a situation of conflict, with several members trying to explain their difficulty in helping beyond the app, especially because of their work routine, childcare responsibilities, or limited financial conditions. The fact that many participants disagreed with Yannine made her leave the group for a few hours. It was up to Mrs. Nora, (oldest member of the group, 64-year-old woman, mother of a 16-year-old daughter; she used to work as a lawyer in Venezuela and now works as an elderly caregiver) to appease the situation. Mrs. Norka developed a unique ability in the group to encourage participants with messages of gratitude, faith, and optimism. In her post, she asked the group to reflect on the content of their messages and reminded them that all participants were going through emotional and financial hardship because of their migratory condition. In her own words: We [the group members] are doing our best in providing support, but I believe this WhatsApp group is not a place "to do therapy" and that many people could be hurt by the level of pressure and judgment that Yannine imposed on us.
These messages were also sent to Yannine in a private chat, which helped the group overcome this conflict and recreate possible meanings of belonging and a sense of community. Through access to this WhatsApp data, the researchers were able to learn how the technology offered a space for the conflict to unfold as well as to resolve it.
Furthermore, conflicts in the group provided researchers with access to information about how participants negotiate appropriate sharing and group norms. Throughout the WhatsApp project, participants shared images and videos of important moments that happened in their lives during settlement, such as finding a job, moving to a different city, birth, finding love, etc. On some occasions, there were disagreements in the group about the type of content that could be posted, as has occurred with an adult-content video featuring nude images of people which was shared by one of the participants. Such content ignited anger in another participant, Karen (43-year-old woman, civil engineer, living with her three children and husband in one of the refugee shelters), who said that this kind of visual material was not suitable for the group and that she feared that her children could see this message, as the phone was shared with her family. Other participants also commented on the ethical aspects of the video, dividing opinions about the freedom/restriction of posts on this theme. Several times, we as researchers did not participate in group discussions due to the uncertainty throughout the project on whether our role as project leaders could have any coercive effect on participants' interactions in the group (Arroz et al., 2019). Such an approach highlights the opportunity of collecting data in ways that might have not been possible due to power differentials that tend to lead to self-censorship among participants. For instance, participants chose not to accept that certain contents (e.g., pornography) are shared in the group while reinforcing the rule regarding group membership after one of the members asked if an acquaintance could be added to the WhatsApp group.
From in situ and real-time observations of participants' interactions enabled by the WhatsApp platform (Kaufmann & Peil, 2020), it was possible to attain deeper knowledge of the conditions of digital precarity experienced by Venezuelans settling in Boa Vista. Throughout the WhatsApp group experience, participants found additional ways to connect their devices to the internet in different city locations. Amelia (41-year-old woman, housewife) said she would often go with her three children to a nearby mall with free internet connection so that everyone in her family could use the phone without having to acquire data, whereas Diego (32-year-old man, resident in a refugee shelter) reported that he accessed the internet in a public square since the shelters did not offer access to communication devices, nor Wi-Fi.
While these examples can be associated with challenges of retaining refugees' participation in research using messaging apps, they also serve to highlight the everyday acts of resistance to economic hardship and exclusions characterizing migrant struggles in diverse settings (Varela-Huerta, 2021). This can be seen in the case of Leonardo (a 26-year-old man who migrated to Brazil to work as a mason) who sold the smartphone device he received from the project to provide financial support to his family in Venezuela. Leonardo shared this information on a confidential basis only with us, researchers (via the private chat on WhatsApp), when making the decision to withdraw from the project.

Weaknesses in Using WhatsApp as a Tool for Data Collection Among Refugee Communities
The analysis of WhatsApp messages showed that the platform provided a space for Venezuelans to interact almost daily during the entire project. Compared to responses to the weekly questions, the methodological dynamics that stood out the most in the WhatsApp group were spontaneous interactions among group members, which allowed researchers to collect data in ways that were not previously planned. These voluntary, unstructured conversations in the group had the advantage of providing rich, unexpected insights into participants' lived experiences and "in real-time" (Kaufmann et al., 2021). Yet, it was important to account for the differences that emerged regarding the engagement of participants in WhatsApp group communication, as this can be a disadvantage when using WhatsApp as a tool for data collection in research with refugee communities. It appears that the level of interaction in the WhatsApp group was related to the kinds of motivations that Venezuelans had to join the group, such as the case of Adriano (48-year-old male, unemployed, and living in a refugee shelter with his wife and three children), the participant who interacted the most with group members. As Adriano explains: "Being part of the WhatsApp group was an opportunity to make friends and enhance my social network." Nonetheless, a small number of participants did not engage constantly, highlighting the challenge of retaining participant interaction in remote research (Kaufmann et al., 2021). Mario (a 31-year-old business administrator, unemployed, father of a newborn), was active in answering the weekly questions of the project, but less engaged in the group conversations. We learned from some participants that the conversation topics and participants' behavior affected their level of engagement in the group. Daniel was a 19-year-old man, resident of one of the city's refugee shelters, who dreamed of becoming a soccer player. When talking about the interactions in the group, Daniel stated that he was not very interested in the subjects that emerged during group chats. For Marta (a 31-year-old single mother with three chil-dren and unemployed), her busy daily routine was not an impediment to participating in the WhatsApp group. She commented that her engagement became increasingly less intense as she felt disappointed with some of her peers: "The attitude of three people in the group is negative; they are proud, arrogant….For this reason, I have not participated much lately. They believe the group is for gossiping." Although the varied engagement of participants in WhatsApp group chats could potentially lead to an imbalanced representation of participants' voices and experiences, the methodology in the WhatsApp group context we studied was not designed to have control over participants' posting behavior (e.g., fixed schedule for posting new content or sending messages). Instead, the possibility that members had to engage in spontaneous interactions in the group offered better opportunities to access different forms of knowledge that were not foreseen but spontaneously brought by participants (Henry et al., 2016).
On the other hand, we recognized from our experience of using WhatsApp in a group context that it was not always possible to follow the discussions as they happen because of the great amount of information shared and the relatively quick time of emergence and resolution of conversations in the group. At least once during the day in our different time zones, we checked the group activity and interacted with participants by answering questions they raised (e.g., about legal procedures, city addresses, etc.), or commenting on the content they shared in the group chat. Similar to other studies Manji et al., 2021), we needed to consider the limitations of data collection and sharing using WhatsApp in research with low socioeconomic status populations. It became clear that some choices when it comes to sharing are related to internet capacity, which may constrain the availability of data for analysis in ways that cannot be predicted. In our WhatsApp study, we observed the employment of digital strategies among participants regarding the data economy, such as avoiding the download of very heavy video content and apps or storing them in their devices to avoid using internet data and searching for free Wi-Fi. This can be illustrated in one of the posts by Mariana (22-year-old, studied architecture in Venezuela) who was careful not to share her videos in the WhatsApp group to prevent group members from spending their data (see Figure 4).
Another important aspect to highlight when using mobile messaging apps as part of a data-gathering strategy among refugees concerns the datafication of these populations (Witteborn, 2021). The process of datafication transforms individuals or populations into digital data that can be traceable and categorized (Adey, 2004). Digital data generated from social media activity (searching for information, navigating through a specific location, or sharing photos online) can be captured by government and private actors to verify refugees' identities and determine their access to rights, assistance and services (Madianou, 2019;Witteborn, 2021

Message 2
This videos are only video presentaƟons.

Message 3
These* from this research that participants engaged in different social media platforms and apps through their smartphones to make sense of their new surroundings and navigate settlement on their own terms. Yet, it is uncertain whether our refugee participants were aware of how their data are collected for the interest of the digital platforms they used, including WhatsApp. While datafication concerns a major challenge to the development of contemporary digital methods, this process does not necessarily make WhatsApp a problematic platform for data collection in refugee contexts. In fact, previous research suggests that the end-to-end encryption introduced in the application to protect user privacy can facilitate data security (Rössler et al., 2018, as cited in Kaufmann et al., 2021. In her study with Syrians in their new city of Vienna, Kaufmann (2018) showed that her participants felt comfortable using WhatsApp to engage in daily chats with the researcher, which helped build trust throughout the process. This is particularly relevant when projects include vulnerable populations.
Finally, concerns regarding the possibility of having the smartphones stolen were widely spread among participants, given the insecurity of life in Boa Vista. In this case, it is recommended that researchers and participants discuss these risks and collectively create guidelines around safety, such as not using smartphones in dangerous areas (Marzi, 2021). For instance, some participants said that they did not take their devices with them when they left the shelters, whereas others reported being cautious about using their phones both inside the shelters and in the city's public spaces. In one case, the participant had his phone stolen and it was only after a friend lent him the phone that he could inform the group that this had happened.

Conclusions and Critical Considerations
In this study, we adopted a longitudinal design using the Conexión Sin Fronteras (Connection Without Borders) WhatsApp group as an intervention project to build community and explore the settlement experiences among Venezuelan refugees in Brazil. We focused on the analysis of participants' interactions and activities in the WhatsApp group while assessing the opportunities and limitations of using WhatsApp as a data-gathering tool in refugee research. In so doing, we demonstrated the potential of WhatsApp to capture refugees' digital and physical spaces of everyday life through its application for data collection in a less structured way. This enabled us to access unique and unexpected occurrences and situations within participants' daily routines and follow important events, changes, and hardships during their lived experiences of adjustment to the new surroundings. We also showed that the experience of accompanying refugees in the WhatsApp group revealed relevant insights into relationship-building in a group context and that went beyond the digital chat, which in turn allowed us to meaningfully engage with the agency of participants to reach out to each other to request and provide help as they resettled.
However, the use of WhatsApp in research with refugee communities does not come without challenges. The lack of engagement in WhatsApp group chats for some refugee members and the impossibility for us to be continuously active during group conversations are factors that researchers need to consider when adopting an unstructured approach to data collection via WhatsApp groups. At the same time, given that the process of collecting data in longitudinal designs can be much time-consuming for both participants and researchers (Kaufmann et al., 2021), flexibility was required when establishing weekly contacts with participants for the questions or conducting daily checks of the activities in the group. Acknowledgment of limitations regarding internet access is key to understanding the data that are produced in projects involving the use of mobile phones in marginalized and vulnerable communities.
In this context, researchers need to account for the existing social and economic inequalities that can affect the adoption of messaging apps in the research process . At a global level, where lowand middle-income countries have become the locus of increasingly intense journeys of refugees (Baeninger, 2018), we believe that a series of precautions are necessary when conducting research in digital environments based on refugees' experiences in south-south routes. Here the questions of "who the access to digital connectivity is for," "where it comes from," and "how it is established" become relevant, especially regarding forcibly displaced populations, located outside of the Western experience. It is in this context that we claim that research cannot serve to further reinforce digital vulnerabilities and inequalities in these locations. Precarious digital infrastructures in some places in the Global South, financial difficulty in acquiring a digital device with constant access to data packages and digital literacy issues can be frustratingly impactful for participants and researchers aiming to use WhatsApp as a research methodology.
Thus, it is important to reflect on geographic locations, social markers, and collective life trajectories, as well as the specificities of refugees' experiences and needs in relation to digital connectivity. Aside from ensuring that the lack of financial means to own a mobile phone device with internet and digital skills levels are not aspects constraining the selection and participation of refugee populations in research, researchers should adopt methodological approaches to help reduce the risk of reproducing power inequalities in refugee research. Taking into consideration participants' motivations to join the research project, including both material benefits and the possibility of access to various forms of support, can foster the development of reciprocal relations between researchers and community members (Restoule, 2008). Through initial interviews as part of the multi-method approach adopted, we could identify participants' needs related to media use, which informed the design and implementation of our WhatsApp-based intervention. At the same time, the opportunity we had to conduct monthly face-to-face meetings with participants during the project supported further reflections about their settlement experiences across online and offline spaces, enhancing their trust and engagement in the WhatsApp group.
Moreover, there is an opportunity for increased digital media literacy that researchers need to meet by engaging participants in discussions about data obtained via mobile phones, apps, or social media. The same opportunity is valid and has already been advocated in other epistemological reflections on the use of digital technologies in migration governance (Madianou, 2019), including critiques of digital apps by development agencies that increase datafication and constrain the production of migrant knowledge (Witteborn, 2021). The intensity of forced displacement movements in the Global South also represents an opportunity to intensify decolonial methodologies in this area of study. A collaborative and spontaneous approach to the use of messaging apps, such as WhatsApp, can help refugees and researchers construct new forms of knowledge production that are inclusive, sustainable, and meaningful in forced migration contexts.