Digitally shaped ethnographic relationships during a global pandemic and beyond

Adapting fieldwork and building rapport in pandemic times

Amid the COVID-19 response from researchers, diverse suggestions have addressed means of altering data collection. For example, Vindrola-Padros et al. (2020) discuss the strengths of employing rapid qualitative methodologies, while highlighting the multiplicity of challenges faced by researchers such as questions of ethics, data collection, and collaboration among multiple investigators or institutions. According to those authors, in efforts to push qualitative methodologies forward and engage with COVID-19 reality, rapid research can take place in “brief” amounts of time, ranging from weeks to months. Meanwhile, some scholars encourage researchers to continue with longitudinal or long-term data collection using alternative methodological frameworks entirely, such as collaborative autoethnography (Roy and Uekusa, 2020) and digital ethnography, to garner rich insights into digital platforms, networks, behaviors, and sociality (Góralska, 2020). As ethnographers explored these alternative methodologies, they promptly created and disseminated crowd-sourced documents listing potential digital methods with links to resources about each method (Lupton, 2020) while “how to” videos introducing digital methods made their way onto Youtube (Vitalities Lab, 2020).

Ethnographers untrained in digital methodologies, however, may be unable to rapidly educate themselves for online research, especially when considering other pressures created in pandemic times. Further, existing research agendas may not be amenable to a shift to ethnography in virtual spaces. The expectation that all researchers can make this rapid methodological transformation also undercuts the extensive training of digital ethnographers, while simultaneously neglecting the potentiality for digital surveillance of research activities (Boellstorff, 2020) and the myriad problems non-digital ethnographers face in taking their work into the virtual world (Valdez and Gubrium, 2020).

For those engaged in fieldwork in primarily offline spaces, scholars also provided advice for making straightforward methodological adjustments to long-term ethnographic fieldwork through utilization of digital technologies such as phone and video conferencing (Lobe et al., 2020; Moises, 2020). Ethnographers have, after all, long incorporated digital technologies such as voice recorders, digital cameras, and computers into traditional fieldwork, which is sometimes referred to as “digitally enhanced” ethnography (Duggan, 2017:5; Hsu, 2014). However, guidance on ethnographic relationships in these contexts is less developed. Here, we examine how to adapt existing ethnographic research that has either already begun or that is being conducted by those who are not digital ethnographers, to digital spaces and technologies. Specifically, we investigate lessons learned in how to build and maintain rapport when access to the fieldsite is limited and these adaptations are necessary. This article posits that for long-term ethnographic studies, research design has to not only consider the nuts and bolts of data collection but must flexibly explore avenues for gaining access, recruiting participants, and building researcher–participant relationships.

In what most consider conventional ethnographic relationships, rapport-building is framed as a process of multiple encounters over time as the researcher and community members interact and build empathy (Edirisingha et al., 2017; Prior, 2018). Time spent in the field contributes to social contact and the development of a working relationship between the researcher and participants. Over time, participants can become comfortable and develop trust with the researcher. Fostering this ease of communication facilitates formal and informal interviews and creates an atmosphere that encourages information-sharing (Le Dantec and Fox, 2015) that extends beyond just data collection by researchers but opens up opportunities for insight into and discussion on how research can be meaningful for participants and their communities.

Challenges in rapport-building, however, have been a part of the ethnographic process since its colonial origins in anthropology and throughout various disciplinary “turns,” with added challenges amid efforts to “study up” (Nader, 1969), conduct research at home, or “otherwise” twist “the gaze” (Springwood and King, 2001). Springwood and King (2001) note that as ethnographers have “(re)inserted themselves in uneasy fields,” such as studies of science, whiteness, and wealth, they have encountered new obstacles as they engage participants more critically. Conventional notions of rapport, for instance, are constructed around ideas of honesty, loyalty, and trust, which can be difficult, if not undesirable, to negotiate and maintain in contexts and multi-sited spaces of exploitation, violence, impossible collaborations, and competing truths (Chakravarty, 2012; Fox and Garcia, 2021; Pitts and Miller-Day, 2007; Sherif, 2001; Springwood and King, 2001; Wood, 2001).

While researcher–participant relationships are considered a “qualifying marker of ethnography,” researchers also contend with dilemmas in rapport maintenance as fieldwork comes to an end (Boccagni, 2011). As Boccagni (2011: 737–738) states, researcher–participant relationships have “traditionally been a matter for secondary concern in academic textwork” as ethnographers detach from field sites. For ethnographers working in disaster and crisis contexts, however, relationships developed in the field may constitute a central priority during the writing phase, especially in situations where researchers experience vicarious trauma as a result of witnessing devastation (Dominey-Howes, 2015) and where the emotional response of the researcher to participants and their situations is considered a valuable avenue for analysis and understanding (Lund, 2012). Scholarship on conventional ethnographic fieldwork in such uneasy fields thus variously addresses relationships and rapport-building between the researcher and participants. When relying on digital tools and digital spaces to augment otherwise traditional in-person ethnography, the concerns outlined above still apply.

To frame these discussions, we draw primarily on ethnographic disaster research experience that has taken place in the southeastern United States, including the 2008 Tennessee Valley Authority (TVA) coal ash spill in east Tennessee, Hurricane Florence in North Carolina, and more recently research conducted in the months following the March 2020 tornado outbreak in middle Tennessee. The tornadoes claimed the lives of 25 individuals, injured 309 survivors, and resulted in over 1.6 billion dollars in damages (NOAA, 2021). Within a week, the COVID-19 pandemic was documented in the United States, limiting fieldwork. Rapport-building and snowball sampling could not take place in person and interviews, networking, and fieldwork relied heavily on digital tools. Noting the challenges to rapport and participant “buy-in” amid disaster research and especially during COVID-19, we first discuss rapport in terms of gaining access to the community and recruiting participants. This is followed by lessons learned in building relationships for and during digital interviews.

Gaining access, establishing a presence, and recruitment from afar

The participatory dimension of fieldwork is typically what distinguishes ethnography within the broader umbrella of qualitative methodologies. Robust ethnographic research, as Harrison (McGranahan et al., 2016) explains, requires researchers to decenter themselves in the real world contexts of everyday lives and, over time, ethnographers have opportunities “to ferret out information from lots of different locations, intersections, and channels.” This participatory process, Boellstorff (2008:65) explains, is not restricted to offline spaces. Researchers can “step into the social frame in which activity takes place,” whether that involves feeling “the heat of the fires” of swidden farmers or slaying “virtual dragons in the company of guidemates.” When ethnographers become familiar faces, or perhaps familiar virtual avatars, people open up and elaborate in ways that may never happen during an interview, emphasizing the importance of ethnographic relationships and rapport-building

Gaining access and establishing a presence is a key prerequisite to relationship building and can be challenging in any context. Pre-COVID-19 challenges in gaining community entry in disaster research already existed. For example, in a non-pandemic context, following the 2008 Tennessee Valley Authority (TVA) coal ash spill, community members impacted by this industrial disaster were simultaneously being approached by environmental groups, TVA representatives, state and national news outlets, and law firms, with different and sometimes competing agendas (Eldridge, 2013, 2015, 2018). In this instance, gaining entry and recruiting participants proved difficult and attempts were sometimes met with immediate disinterest. The challenges escalated as the disaster became increasingly politicized and imbued with lawsuits. In those cases, some people were interested in talking about their experiences, but not interested in participating in a more formal, semi-structured interview. Fortunately, because ethnography is flexible and adaptive, fieldwork became focused primarily on public meetings and events, informal conversations, and open online forums, rather than semi-structured interviews (Eldridge, 2013:10–13). However, this highlights instances in which ethnographic relationship building can be challenging even when face-to-face interactions are available.

Typically, upon entry into the research context or community, the researcher can begin establishing a presence by attending events, meetings, and daily activities that assist in identifying potential participants (Horsely, 2012). During research for Hurricane Florence, for example, fieldwork regularly involved attendance at long-term recovery group (LTRG) meetings and town hall assemblies (Reinke and Eldridge, 2020). LTRG meetings began with introductions, which not only provided an opportunity for researchers to explain their research to the recovery group members at once, but also allowed researchers to identify and learn about representatives from diverse organizations involved in recovery work. Town hall meetings additionally provided opportunities to interact with survivors and observe interactions among survivors, recovery organizations, and public officials. These approaches often fit collaborative research frameworks that prioritize partnership with organizations, research participants, or other community entities throughout the research process (Reinke, 2019). Ethnography at a distance does not afford for these types of interactions, creating a need to rethink ethnographic relationship building as COVID-19 responses throughout the United States shifted the nature of public gatherings, moving many to virtual platforms.

Fewer opportunities for face-to-face encounters hampered traditional snowball sampling in the field which relies on good rapport and referrals made by participants. In lieu of these in-person opportunities, digital forums can be scoured for insight into and connections to one’s research site, as well as starting points for identifying potential stakeholders. This approach of course will vary depending on one’s focus, but following meaningful hashtags (Bonilla and Rosa, 2015) and seeking out social media pages for elected leaders, public officials, and local interests groups in places such as Facebook, for example, could reveal concerns, debates, events, and other valuable information. While there are debates about what constitutes “participation” in online research and differing views about whether or not viewing asynchronous interactions in digital spaces is “observational” versus “engaged listening” (Praechter, 2012), we consider our approach in these contexts as observational, specifically reviewing information posted in publically available online spaces.

Specifically following a disaster, social media pages often arise for emergency management purposes, to facilitate communication, and provide platforms for organizing, fund raising, and maintaining connections (Finau et al., 2018; González, 2020; Lovari and Bowen, 2020). On these pages, members often reference involved organizations, community leaders, impacted businesses, volunteer opportunities, public meetings, relief funding information, and others who could be potential contributors for purposive sampling. A prime example of this can be seen in the aftermath of the 2020 tornadoes in Tennessee. Public officials, in collaboration with the local television station in one semi-rural town, utilized their Facebook and Youtube accounts to provide updates to the community. Initially these were daily briefs as dozens of people were still unaccounted for, but became periodic occurrences once COVID-19 became a reality. Another social media group was created for community members to relay information about funding, volunteering, lost and recovered personal belongings, housing, and other pertinent information. These were all publicly open pages and groups where we could browse comments, follow events, identify stakeholders, and glean initial insights into local concerns from a distance.

Digital ethnographers may be best suited to offer advice into thoughtful analysis of virtual spaces (Góralska, 2020). Some scholars, for example, have pursued online “ethnography at a distance” without any contact or engagement with research subjects (Forsey et al., 2015), while others are more interactive, engaging in online virtual worlds (Boellstorff, 2008; Snodgrass, 2016) and focus groups (Stewart and Williams, 2005). In terms of recruitment, there is scholarship addressing ethical considerations for approaching participants via active and passive recruitment on social media (Gelinas et al., 2017; Hugelius et al., 2017). Edirisingha et al. (2017), for instance, suggest using social media such as Facebook to foster friendships with participants and minimize the power hierarchy between researcher and participant. Here, we are not suggesting a major shift into a new-to-you form of research design. Rather, using the digital platforms available—an approach that can enhance in-person fieldwork as well—could assist in the initial process of identifying potential participants. However, in the context of disasters, survivors may be inundated with social media information, scammers, or journalists, confounding the positionality and goals of the researcher. To ensure transparency and clarity, we found the use of more professional lines of communication, such as using university email accounts, to be more effective for eliciting responses after potential participants were identified. Given the reduced in person interaction, this can assist in establishing a more trusted presence in the community rather than appearing as an entity working solely through social media platforms.

After the Tennessee tornadoes, to illustrate, the community quickly became overwhelmed with thousands of volunteers, insurance agents, lawyers, emergency response personnel, disaster tourists, and civil society organizations. The vast array of interlocutors in both offline and online spaces made it difficult for organizations and survivors to delineate legitimate volunteers and relief workers from scams and con artists. During research on tornado preparedness and response, one communications coordinator working primarily in social media commented, “I had strategically picked somebody to be one of my (information) monitors whose dad was at the Emergency Operations Center so that I could make sure that the news coming in and out was correct and that I was sharing the right stuff—because in the first few days we had some really big problems with that. We had scams, and we had all of that stuff in there pretty quickly” (Interview, December 2020). With enthusiasm, the interviewee continued to explain that in one case a woman “said that she was making quilts and was taking money for survivors to … make them a quilt and they never saw those” (Interview, December 2020). This illustrates how in some contexts, there may understandably be skepticism when one identifies themself as a researcher in social media spaces attempting to recruit participants.

Concerning access to survivors, engagement with community leaders, who function in many ways as gatekeepers, was pivotal and required time and a delicate approach. Given that the research is disaster-based, this poses a number of ethical dilemmas for emergency responders and community organizers who engage with survivors and researchers, which range from data security to hospital response management (Hayes and Kelly, 2018; Sanfilippo et al., 2020; Veenema et al., 2017). With an understanding that trust and social support in leadership influence survivor responses and recovery (Eadie and Su, 2018; Han et al., 2020), it was clear that the grassroots community leaders had fostered trust with survivors and put great effort into streamlining contact with survivors. This was to minimize the barrage of individuals approaching survivors—such as those seen in the TVA coal ash spill—and to create security regarding survivor privacy. This important protective relationship between community organizers and survivors also created snowball sampling challenges for research recruitment. While describing recovery group relationships with survivors, one recovery worker noted “they have gotten extremely sensitive about survivor information” (Interview, January 2021). The interviewee referred our team to other recovery group members with more direct connections to survivors.

These navigational challenges to gain access and build relationships, however, are not unlike gatekeeping issues that may be seen in diverse cultural contexts. For example, in human rights related research with the Acholi of Northern Uganda, researchers relied on both local government officials’ and clan chiefs’ approvals to conduct interviews in the region (Kim et al., 2018b). Similar hesitation has been seen in research conducted with case managers for human trafficking survivors who were unsure how the confidentiality of research would intersect with the confidentiality of case management. While e stablishing a presence and building rapport in person assisted in overcoming that obstacle upon field entry in the human trafficking case (Kim, 2018a; Kim et al., 2018b), in Tennessee, we were able to make headway into the tornado affected community from afar with time and through pointed efforts to build trust with community leaders.

Because leaders had established secure communications networks with survivors, our team also requested that information about our research be disseminated by leaders through those networks to assist in developing community awareness regarding our project since our physical presence would not be a visible indicator in the community. Exploring social media, local websites, and other virtual spaces may not provide full access to key participants, but it at least may provide researchers with starting points for identifying stakeholder and avenues for trust-building and entry through gatekeepers.

Building relationships and rapport during virtual interviews

Once key participants are identified and initial connections are made, how, then, do ethnographers build and maintain relationships with participants in virtual contexts where interactions with participants are limited primarily to the interview time itself (Fox and Garcia, 2021; Weller, 2017)? Strategies developed to address this question have implications for research in both pandemic and post-pandemic times (Fox and Garcia 2021; Gray et al., 2020). Such strategies help researchers find valuable ways for expanding their methodological repertoire in recognition of “the multifarious ways in which ‘the digital’ is now intertwined with everyday life” (Duggan, 2017:8), which has arguably been illuminated during the current pandemic.

There are obvious limitations to key functions such as rapport-building when relying primarily on digital interviewing, as Weller’s (2017) study highlights, such as disruptions from poor connection quality or lack of auditory clarity, which cause frustration and hinder relationship building and maintenance. Lack of access to the internet in rural areas, digital security, and other digital gaps reflect broader systemic issues that can limit participation in research interviews (Irani, 2019; Moyle et al., 2020; Pollard and Jacobsen, 2019; Ruiz-Martínez and Esparcia, 2020). Consideration for interview location is an issue in any format or context. Interviewees connecting virtually from home or work may have distractions, but there can also be many perks for both researchers and participants to virtual interviewing, such as the flexibility of location and the ability to remain in the comfort of one’s own personal space, as opposed to the “pressure of presence” that may be felt in face-to-face settings (Weller, 2017: 618).

After losing most of the in-person component of ethnographic research in Tennessee due to pandemic restrictions, our team focused on building relationships through the virtual interviewing process. Each step of the way, we incorporated some strategies that facilitated rapport-building in the context of a compounded disaster. For example, Deakin and Wakefield (2014: 610) discuss the use of pre-interview email exchange to facilitate the rapport-building process. While researchers often have to provide recruitment email templates per their Institutional Review Board requirements, customization of those scripts can be useful for creating a less generic and more inviting tone in initial emails. In the context of the March 2020 tornados, pre-interview email exchanges often included introductions, gauging interest in the project, scheduling interviews, sharing information or links about our work and previous research, and sending documents, such as consent forms that can be reviewed prior to more thorough discussions on consent that ensue at the start of virtual meetings. Deakins and Wakefield (2014: 613) note that discussion of consent at the start of an interview can set a more formal, uncomfortable tone; we therefore found that emailing the consent form to participants ahead of the interview mitigated some of the formal awkwardness often associated with consent processes. Participants had a chance to review it in advance and had a familiarity with the document prior to it being discussed. This gesture also assisted in demonstrating efforts to be aware of the needs of participants and assuaged pressure some may feel when presented with a formal document—especially having been inundated with legalese via insurance agencies, FEMA, and other venues of aid.

In situations where there was a significant time gap between initial contact and the interview date, participants were told they would receive a link for the teleconference a day or two before the interview, which also served as a reminder. Telephone calls or even text messages were sometimes interjected as modes of maintaining communication. In several cases, we found that participants had taken advantage of links we provided to our university or personal websites for vetting purposes and for learning more about our disciplines and research interests. Some participants even referenced our publications during interviews. Disaster responders were particularly interested in other field studies and often came to the interview or preceded the interview with questions for discussion. Pre-interview communications can thus serve to link participants to previous work, providing a sense of who the researchers are and using public information to reinforce authenticity.

At the start of each interview, we first explain the project and carry out the consent process prior to audio recording. We spend time introducing ourselves and talking about what we do and why with a casual tone as everyone adjusts to the interview platform and settles in for deep discussion. While pre-interview conversations are commonly used to put participants at ease in any interview setting (Pinksy, 2015: 285), in virtual spaces, we also take a moment to create anecdotal banter while describing technological setups. For instance, at the start of one interview, Reinke references her “command center” referring to the multiple monitor displays in a setup similar to the participants to ease introductory awkwardness. This is also done in part because notes may be typed on a separate monitor, and given the COVID-driven increase of virtual meetings that have questionable levels of participant attention, the author made sure to communicate to interviewees that she was still paying attention while looking elsewhere for note taking. While in an in-person scenario this would be apparent, it was necessarily clarified to prevent damaging rapport.

Ethnographers are already familiar with human subjects research protocols and must adhere to protocols and standards regarding informed consent, voluntariness, and minimal harm, although these protocols entail necessary shifts during the pandemic (Fox and Garcia 2021). While consent forms—whether digitally documented and signed prior to your meeting or verbally provided at the start of an interview—explain the sensitivity of questions and potential for discomfort, gauging comfort and interest may be challenging via telephone or video interviews where visual cues are either absent or less discernible than in person. It is therefore not only important for researchers to continuously and actively listen and watch for cues or expressions of discomfort or disinterest but to also take opportunities to pause and remind participants that they do not have to continue with a particular line of inquiry.

In our disaster research, for example, where conducting interviews with survivors, volunteers, and response and recovery workers in disaster aftermaths can trigger emotional responses, continual assessment of the interview is necessary to avoid exacerbating participants’ extant negative emotions or trauma. According to one interviewee we spoke with 3 months after the tornadoes, “it’s still difficult to talk about” but “you don't realize whether you're ready or not until you start talking about it” (Interview, June 2020). Maintaining mindfulness and flexibility in the interview process and a willingness to discursively shift directions is important for building and sustaining rapport, particularly with interviewees who may need to maintain boundaries and barriers around particularly triggering topics. This is applicable broadly to research amid the stress of the pandemic—a global health crisis requiring ethnographers to be “extremely alert, self-reflexive, and sensitive” (Ghosh 2020).

Research has shown that creating a comfortable interaction during interviewing improves memory recall ability of participants and correlates to lower reports of misinformation (Vallano and Compo, 2011). Acknowledging this, prior to interviews, informal or non-project related information about participants was noted that could assist in the interview dynamic. For example, when interviewing a business owner impacted by the tornado, Eldridge lightened the mood after a lengthy discussion on tornado destruction by asking how the business’ mascot was doing, which elicited a laugh from the interviewee, who had publicly posted a picture of the mascot as a survivor amid the tornado destruction. As the pandemic led to greater familiarity with video calls, individuals appeared to have relatively little trouble engaging with videoconferencing technologies, although some elected to use phone calls.

Although traditional ethnographic fieldwork might conclude an interview with a cup of coffee or a beer, a chat while walking to an event, or an otherwise more normative and informal interaction, digital interviews come with no such possibilities. The interview, and the rich and difficult conversations it often envelops, simply ends; yet as Weller (2017: 617) explains, “‘leavings’ are as important in rapport as ‘greetings’ in terms of fostering a long-term connection.” Ending the interview with a tangential question may ease discomfort, open doors for continued communication and participation, and conclude in a way that makes the participant feel that there are actionable opportunities. Frequently, toward the end of our interviews, participants were asked to share if they had positive outcomes or lessons learned from both the tornadoes in Tennessee and the COVID-19 pandemic. Responses varied widely, but participants often commented on how pleased and proud they were of their community’s solidarity, willingness to work together, and community spirit. Shifting from individual experience of devastation and pain to a collective notion of support proved a positive end to interviews. In addition, during some interviews concerning the Tennessee tornadoes, COVID-19 interestingly provided a less emotional, mutually relatable topic for participants and researchers to discuss, creating opportunities to shift the conversation while creating empathy. For example, at the closure of one interview Eldridge mentioned that they “hope to visit again; it’s kind of hard with COVID” (Interview, January 2021) and the interviewee and Eldridge immediately struck up a conversation about the area, their ties to the region, and potentiality for future visits. While the lines of inquiry and the content of banter will vary depending on one’s research topic, the point here is to highlight the need for developing comfortable interview exit strategies specifically designed for virtual contexts.

With long-term ethnographic research, continued follow-up conversations with key participants is also pivotal to maintaining relationships from a distance. In our research, scheduling casual videocalls with emergency managers, for example, not only allows us to learn about ongoing recovery issues, but also communicates to participants our commitment to research in their community since our physical absence does not afford visible reminders. Even from afar, follow-up communication can provide opportunities for what Pitts and Miller-Day (2007) refers to as positive “turning points,” characterized by deepened relationships and increased familiarity and comfort between researchers and participants.

Finally, of note, during traditional in person fieldwork, researchers become intimately familiar with landscapes and the built environment, and these become points of commonality between the researcher and participant during the interview processes as locations are referenced. Eldridge has intimate familiarity with our Tennessee field sites, for example, and was able to connect with our virtual participants about local schools, community sites, and roads. Using geospatial information about the built environment impacted by the tornados assisted in relationship building during the interview process. Our team used Google Earth, Google Maps, Google Images, and news reports, in addition to extant knowledge of the area, to create a visual familiarity with areas that were being referenced by interviewees. Use of such technologies to develop familiarity of landscapes can be useful in conventional fieldwork as well. In the wake of a disaster, for example, physical access to affected areas may be reserved for emergency response personnel, utility companies, and survivors. This can hinder a field researcher’s ability to become familiar with the physical landscape. Whether doing ethnography from a distance or in constrained fieldwork situations, having shared spatial familiarity was key for fostering discussion as participants referenced relevant locations. While the strategies outlined above may only seem like small adjustments, we found that consistent implementation significantly facilitated rapport-building and such strategies can quickly be applied in situations where conventional fieldwork plans are unexpectedly interrupted.