The COVID-19 pandemic has instigated a profound reconsideration of how we understand fieldwork. Our new “socially distanced” reality brought about by the pandemic has “pushed us back into the armchair—both in a physical and metaphorical sense—and required us to utilize new methods to conduct research from our own homes” (Howlett Reference Howlett2021, 12). Instead of travelling to, or embedding ourselves within, particular “field” sites, many researchers have come to use telecommunications platforms like Skype, Zoom, Microsoft Teams, and WhatsApp to conduct interviews and focus groups. Some scholars have immersed themselves in digital communities like Reddit, Facebook, Twitter, and Instagram for ethnographic research, or “net-nographies,” while others have increasingly collected archival materials from digital archives and audio-visual data from social media platforms like YouTube and TikTok (Hanson et al. Reference Hanson2011; Hartelius Reference Hartelius2020; Hine Reference Hine2000, Reference Hine2005, Reference Hine2015; Shogan Reference Shogan2010; Zeng and Abidin Reference Zeng and Abidin2021). In this way, digital research methods have enabled many researchers to adapt and continue their work in real-time with audio, visual, and textual material without physically leaving their homes (Lobe, Morgan, and Hoffman Reference Lobe, Morgan and Hoffman2020). Although several of these approaches pre-date the COVID-19 pandemic, and much non-qualitative research has been conducted remotely for years, international travel bans, local lockdowns, and social distancing measures have further validated and even provided renewed opportunities for their use. As a consequence, our understandings of fieldwork have generally expanded to include the mundane and everyday practices involved in our work-from-home lives, such as interacting through digital technologies, writing and responding to e-mails, and “liking,” sharing, and viewing social media content (Bluteau Reference Bluteau2019; Pink Reference Pink and Amit1999; de Seta Reference de Seta2020; Howlett Reference Howlett2021; Krause et al. Reference Krause2021; MacLean et al. Reference MacLean, Rahman, Turner and Corbett2020; Will, Becker, and Weigand Reference Will, Becker and Weigand2020).
While scholars have widely written about the challenges, best practices, and lessons learned from their experiences conducting research remotely, the ethical considerations surrounding the use of these methodological approaches still remain underexplored (MacLean et al. Reference MacLean, Rahman, Turner and Corbett2020). Although the merits and possibilities of remote and digital methods can be seen, their usage prompts new ethical questions in fundamentally obfuscating where our responsibilities to research participants and sites begin and end. In mediating what would otherwise be face-to-face engagements, digital platforms change the nature of researcher-participant interactions and merge our “homes” and “fields” in ways not previously conceived. Transparency and replicability in the collection, publication, and dissemination of our data are likewise thwarted when we have minimally, if ever, physically accessed our field sites or interacted with our research participants in person. Moreover, dynamic and politically unstable field sites provoke new ethical quandaries around participant engagement and researcher responsibility, especially immediately after we have left these places, but equally when collecting data from a distance (Knott Reference Knott2019). Given the enduring disruptions imposed by the COVID-19 pandemic, questioning our ethical responsibilities as researchers when conducting digital and remote research—both when continuing projects and starting new ones from afar—has become all the more important since early 2020. As we have now passed the critical juncture we were approaching prior to the pandemic regarding our use of remote and digital techniques for political science research, a mainstream discussion around ethics is necessary.
We seek to encourage scholars of political science to more explicitly query when our ethical responsibilities in research begin and end, particularly when we do not spatially, temporally, or even emotionally embed ourselves within the societies we study. Such considerations are critical given the enduring structural changes imposed by COVID-19 as accessing field sites, archival records, and research participants will remain challenging where entry protocols have permanently changed, vaccine access or take-up is low, testing is limited or unavailable, and viral transmission continues to pose a health and safety risk to populations globally.Footnote 1 A discussion about ethics is also relevant beyond the context of the pandemic and the increased usage of digital and remote research methods, as digital technologies and telecommunications software proved advantageous for the conceptualization and conduct of all types of research before 2020, and will likely only become more beneficial into the future. As these tools offer renewed methodological opportunities and are now part of how we routinely conduct political research, reconsidering our ethical responsibilities is critical for the discipline.
We begin by exploring the ethics of fieldwork as understood before the COVID-19 pandemic, focusing on how political scientists have defined ethics when engaging with research participants both within and beyond the field. We then detail how the “form, degree, and loci” of potential harm (Knott Reference Knott2019, 142) has expanded in light of the pandemic and other crises, and encouraged an increased use and greater recognition of digital and remote methods. Following from here, we highlight why we, as researchers, must more thoroughly consider our ethical responsibilities when conducting research remotely, especially in terms of participant recruitment and access, informed consent, and the publication and dissemination of our findings. We conclude with a discussion of best practices for remote fieldwork when returning (physically or virtually) to field sites we had visited before the pandemic, or entering fields for the first time. While integral to qualitative research, this discussion is equally valuable for other methodologies due to the significance of digital and remote tools for data collection in political science. Our main aim in this paper is to start a larger conversation about our ethical obligations as scholars when conducting research away from the societies and socio-political phenomena we study.
The Ethics of Fieldwork as We Know It
Fieldwork, or field research, is a longstanding method for qualitative data collection used in many social science disciplines, including anthropology, sociology, and political science. While definitions vary across sub-specialties, it is typically understood as a methodological technique wherein researchers physically travel to and locate themselves within a context or field site away from their homes to “acquire data, information, or insights that significantly inform one’s research” (Kapiszewski, MacLean, and Read Reference Kapiszewski, MacLean and Read2015). A foreign context is not necessarily required. Total immersion within a field site or in-person interactions with human participants may not always be the primary goal, or even possible, for some studies. The empirical value of conducting fieldwork is thus in its objective of “gaining firsthand knowledge” (Shaffir and Stebbins Reference Shaffir and Stebbins1991, 5) through the “gathering of evidence in context” (Kapiszewski, MacLean, and Read Reference Kapiszewski, MacLean and Read2015, 7). Because fieldwork involves extensive engagement and interaction with people and materials “in their own setting[s]” (Wood Reference Wood2007, 123), it consequently produces deep insights and rigorous, serious, and meaningful research, especially when compared to previous “armchair” approaches where scholars based their theories and conclusions on others’ research without necessarily directly interacting with the studied populations (Howlett Reference Howlett2021, 3). Since fieldwork of any kind uses a set of dynamic and iterative practices, it often involves multiple trips to and from the site(s) under study and often includes numerous interactions with the people participating in the research throughout the lifecycle of a project (Kapiszewski, MacLean, and Read Reference Kapiszewski, MacLean and Read2018).
As fieldwork in many cases involves directly engaging with people in their unique settings, this methodological approach requires significant ethical considerations to ensure that research is conducted sensitively and harm is mitigated. Discussions have appropriately been ongoing in literature both within and outside the discipline of political science to highlight our ethical responsibilities as scholars. Lauren MacLean et. al. underscore these in their 2018 reflection on research ethics and human subjects, stating that the “foremost ethical obligation and therefore the first duty of scholars is the ethical treatment of people affected by our research, particularly its human subjects” (see Jacobs et al. Reference Jacobs2021 QRD Working Group I.2, 1, online appendix). Although ethical questions and concerns are now intrinsic to all types of research, discussions around research ethics were limited until the introduction of several codes in the twentieth century, such as the Nuremberg Code (1947), the Declaration of Helsinki (1964), and the Belmont Report (1979). The latter code outlined guiding ethical principles for research with humans, focusing specifically on respect, beneficence, and justice to ensure participants engage in research as independent, informed, and consenting individuals treated with dignity and respect (Murphy and Dingwall Reference Murphy, Dingwall and Atkinson2001). To this day, respect for all persons involved is central to any study, beneficence in terms of maximizing benefits and minimizing harm to participants, and fairness in the distribution of “benefits and burdens” associated with the creation and dissemination the findings (Kapiszewski and Wood Reference Kapiszewski and Wood2021). Within the social sciences, scholars are now routinely required to outline and assess the risks their projects may pose in order to be granted approval from their academic institutions, the relevant Institutional Review Boards (IRBs), and also funding bodies. In particular, scholars must explicitly demonstrate that the magnitude and probability of harm or discomforts for participants anticipated in their research are not any more significant than those “ordinarily encountered in daily life or during the performance of routine physical or psychological examinations or tests.”Footnote 2 These risks must then be communicated to participants and informed consent—generally defined as “the right for research participants to make independent, informed, and voluntary decisions about being included in a research project”—is obtained in order to ensure the ethical conduct of research.Footnote 3 While challenges undoubtedly exist in identifying all possible risks in a given study, especially when it involves entering a new setting or field site, these procedures protect participants from potential harm and trauma.
Given the nature of IRBs and their focus on institutional liability concerning researcher conduct, the protection of scholars and study participants from a legal standpoint is thus often the principal concern. As a result, social scientists have often disproportionately focused on the ethical considerations highlighted in IRB and ethics protocols—such as when interacting directly with participants in the field or actively collecting data using experimental manipulations—leading to limited discussions of ethics beyond the training offered by IRBs and the checklists of ethical issues related to the use of specific research designs and methods in certain contexts (Fujii Reference Fujii2012; Guillemin Reference Guillemin and Gillam2004). This focus on “procedural ethics” in both training and institutional review has, ultimately, led to the equation of ethical concerns and IRB approvals while overlooking “ethics in practice” (Fujii Reference Fujii2012, 717). Although discussions around practical ethical quandaries and researchers’ responsibilities, including the ways to mitigate and navigate them, are perhaps most thoroughly present in field research guides, these pieces often overly emphasize objectivity and relationships with research participants while in the field, giving less attention to the ethics of our approaches to research more broadly (MacLean Reference MacLean2006). As political scientists regularly work on topics connected to trauma, violence, and conflict, discussions around ethics in the discipline have also primarily centred around sensitive topics, such as race, gender, and identity, or field experiments and research in “dangerous” sites like conflict zones or unstable political climates (Cronin-Furman and Lake Reference Cronin-Furman and Lake2018; Loyle and Simoni Reference Loyle and Simoni2017; Lyall Reference Lyall, Bennett and Checkel2015; Wood Reference Wood2006). Consequently, discussions about the ethics in the conduct of political research more broadly have been nominal.
In response to these trends, the Qualitative Transparency Deliberations (QTD), sponsored by the Qualitative and Multi-Method Research (QMMR) division in the American Political Science Association (APSA), brought the discussion of ethics and qualitative research to the forefront of the discipline. The QTD deliberations concluded in 2018 with a summary article published in Perspectives on Politics in 2021 and extended reports published online by each working group (Jacobs et al. Reference Jacobs2021). All throughout these working group reports, “ethics in practice” was emphasized, as researchers highlighted the central ethical concern of protecting study participants, in addition to concerns of academic integrity and competence, when approaching a study. MacLean et. al. in particular note that “conducting ethical research goes well beyond the IRB review process and the Belmont principles”; they underscore that informed consent may be challenging to obtain in situations where participants are impoverished and seek compensation or where access to local populations requires consent from institutional or governmental authorities, as confidentiality may be compromised (Jacobs et al. Reference Jacobs2021 QTD Working Group I.2, 4). Attempts to ensure transparency in the research process can also pose particular ethical challenges to “entire communities of political scientists,” as identifiable features about participants in interviews, archival materials, and field notes may come with risks of imprisonment, violence, retribution, or even death (Jacobs et al. Reference Jacobs2021 QTD Working Group I.2, 4-6). Given the sensitivity involved in data collection, especially in light of the new risks to data security that have emerged with technological advancements, a much greater emphasis is now placed on data management and protection plans for upholding ethical research processes across the discipline, and especially in the context of high-risk research (Bond, Lake, and Parkinson Reference Bond, Lake and Parkinson2020).
Although political scientists are now paying much more attention to these data protection policies and data management plans, Eleanor Knott argues that our ethical responsibilities are not limited to our data collection processes, nor do they simply end when we have “exited” the field (Knott Reference Knott2019). In pointing to her own experiences conducting field research in Crimea prior to its annexation in 2014, she contends that field sites are dynamic, and our ethical obligations to our research participants are not contingent on the time and place wherein we conducted research. While Knott’s situation may be particularly niche, socio-political changes in field sites regularly occur, such as in the form of political upheavals, economic crises, and environmental disasters (Cronin-Furman and Lake Reference Cronin-Furman and Lake2018). Events of smaller magnitude can likewise have fundamental impacts on our participants’ lives, introduce new harms, alter the contexts of our studies, and eliminate the key institutions, actors, or behaviors behind the outcomes we seek to explain. As such, our relationships with participants and our fields may integrally change from the time when we collect data to when we present and disseminate our findings. Returning to the same sites is also not always feasible due to dynamic grassroots situations, and because the places we left may no longer exist as they did previously, if they do at all. Whereas ethical concerns begin at project conception, our ethical responsibilities as political scientists clearly extend beyond IRB approvals and continue through data collection, analysis, publication, and replication.
Expanding the Form, Degree, and Loci of “Harm”
Given mitigating harm is a fundamental principle of fieldwork, conducting ethical research accordingly requires a recognition of the non-static and evolving reality of our field sites and homes, as well as a continuous reconsideration of how we define “harm.” While conflict and political and social upheavals have always presented risks to participants and fieldworkers, and therefore shaped the ways we define harm for ourselves and our participants, Carayannis and Bolin (Reference Carayannis and Bolin2020) add that disease poses an additional layer of risk. Although not widely discussed until March 2020, Kanisha Bond et al. (Reference Bond, Lake and Parkinson2020) argue a similar point in their contribution to the multi-party reflection series on “COVID-19 and the Social Sciences” sponsored by the Social Science Research Council (SSRC). In their piece, they purport that the pandemic, and crises more generally, exacerbate existing vulnerabilities, particularly for those societies already facing inequalities in terms of income, class, or identity (Bond, Lake, and Parkinson Reference Bond, Lake and Parkinson2020). In posing a biological risk to any and all people, including both participants and scholars, COVID-19 has consequently shown the potentially significant and life-threatening risks to health and wellbeing posed by viruses and other illnesses. Even though researchers were never required to obtain a flu shot or limit research for fear of infection during flu seasons, the structural changes imposed by the COVID-19 pandemic have fundamentally transformed our colloquial understandings of harm by highlighting that “danger” can take more than physical and psychological forms.
In addition to the health consequences of contracting the virus, the risk of exposure and spread of disease is further increased as researchers travel into and out of their field sites. As the risk of transmission of diseases like coronavirus are highest when individuals are in close physical proximity, traveling to a field site to conduct research—whether located within the same country or abroad—intrinsically presents a risk of exposure to both our participants and ourselves, as well as the people within our direct social “bubbles” and larger communities. In fundamentally placing both researched communities and scholars’ own health at risk, these movements are particularly problematic for regions where levels of transmission are high or uncontrolled, and societies with limited or challenging access to health care professionals, testing, and vaccines. The reality that the virus disproportionately, yet randomly, affects people is also a noteworthy concern, as the risks to any one person’s health cannot definitively be identified. The possibility that some individuals are asymptomatic carriers of COVID-19, just like many other diseases, introduces an additional risk that had not previously been given significant attention when assessing the health risks involved in field research, or even in our interactions with other people in our everyday lives. It is therefore not just a function of where and how often we travel that now involves health and safety considerations, as “dangerous field sites” are no longer the only places with high levels of risk. While not explicitly referencing global pandemics, Knott’s argument to expand our understandings of “harm” in terms of “form, degree, and loci” is hence particularly relevant within the context of COVID-19 (Knott Reference Knott2019).
Although strategies to mitigate risks in the field—both related to COVID-19 and otherwise—have been suggested and proven effective by other scholars, such as hiring research assistants, local enumerators, or personal contacts to collect data on behalf of the researcher, these approaches may only serve to protect scholars, while still posing the same risks to the health of our participants, their communities, and the larger societies we study (Krause et al. Reference Krause2021; MacLean et al. Reference MacLean, Rahman, Turner and Corbett2020).Footnote 4 Since we, as researchers, are responsible for considering how our studies pose risks to both those participating in our studies and ourselves, we must therefore evaluate and defend how accessing a certain field for research purposes is necessary and poses the minimum level of risk possible to a much larger population during and after our data collection. As this has proven exceedingly challenging since March 2020, many scholars have appropriately (re)adapted their own projects to conduct research away from their participants and field sites, calling for the use of remote and digital tools to conduct virtual interviews, archival research, and digital ethnographies (Kapiszewski and Wood Reference Kapiszewski and Wood2021; Krause et al. Reference Krause2021; MacLean et al. Reference MacLean, Rahman, Turner and Corbett2020; Will, Becker, and Weigand Reference Will, Becker and Weigand2020). While these tools pre-date COVID-19, their usage has allowed researchers to continue pre-existing data collection in the face of crises, and has received a broader validation across the social sciences as generally appropriate research methods. As such, the concept of fieldwork has come to be understood as much more than “face to face” data collection, but also inclusive of field sites accessed from home.
While the COVID-19 pandemic has prompted widespread reflection around the adoption and effective use of digital and remote tools for political science research, a discussion about how we conceptualize and understand ethics within this context has yet to be seen. Although ethical considerations are especially prescient for field researchers in direct contact with societies under study, like Knott (Reference Knott2019), we argue that they are equally essential when we are not physically present with our participants or in our field sites. Beyond the case of dynamic socio-political contexts and events, ethical concerns also arise when research is conducted away from our participants and field sites using remote and digital methods. Although digital tools indeed aid in the conduct of research from afar and can mitigate some context-specific harms to both participants and scholars, as was demonstrated from the beginning of the COVID-19 pandemic, “vulnerability can emerge in any space,” whether in person or digitally (Bond, Lake, and Parkinson Reference Bond, Lake and Parkinson2020). For this reason, it remains important to consider ethics and the potential for harm in all contexts wherein we conduct research, particularly to ensure respect, beneficence, and justice. The remainder of this article therefore begins a larger conversation around the ethics of conducting political science research using remote and digital methods, specifically around the additional challenges posed for participant recruitment, obtaining informed consent, and ensuring transparency in the research process and dissemination of our findings.
The Ethics of Remote and Digital Research
For the purposes of this paper, we differentiate between remote and digital field research methods. We define remote methods as techniques used to collect and analyze data for a project that could be collected in person. This includes the use of telecommunications software or applications to conduct virtual interviews or focus groups, or online finding aids and archival records. By contrast, digital fieldwork involves the use of digital data sources that are not a copy of some physical record (tweets from Twitter, posts on Facebook etc.), or embedding oneself directly in virtual communities (for example, ethnographic research in Reddit forums or participant observation on Twitter). As data collection must continue in the face of obstacles to access, as is the nature of any academic discipline, it must be recognized that translating in-person research is not simply substituted with remote data collection; any method comes with its own practical and ethical challenges. Whether adopted to adapt a project in the face of a crisis, or deployed as the primary research tool, researchers relying on remote fieldwork and digital data sources must think critically about where our ethical responsibilities to research participants and field sites begin and end.
Participant Recruitment and Data Access
When engaging in remote fieldwork, and especially when transitioning from in-person to remote data collection, scholars must think beyond what is gained and lost in terms of physically accessing the field. One of the primary benefits of fieldwork chiefly stems from the direct connections researchers can establish with their participants in the field. Often these emerge through organic means, such as networking, canvasing, or informational interviews. Yet, participant access and recruitment differ substantially when connecting with individuals through inorganic means like cold e-mails or recruitment ads on social media platforms like Facebook and Instagram. Importantly, remote recruitment via text messaging can be more productive than recruitment through social media; Mark Reñosa et. al. note that relying on mobile providers to snowball sample in some countries, such as Uganda and Zambia, can reduce issues of access (Reñosa et al. Reference Reñosa2021). Still, studies have often found significant differences in key traits like age, gender, race and political orientation when recruiting remotely relative to in-person interception, especially for hard to reach populations (Guillory et al. Reference Guillory2018). For specific studies, this may generate an over- or under-representation of certain groups in ways that may significantly impact the results, such as by over-representing the outcome of interest or under-representing key but hard to reach populations (Geddes Reference Geddes1990).
Similar limitations are connected to Internet access. As remote fieldwork requires reliable Internet connections, platform compatibility, and technological competence and familiarity by both parties, it cannot be assumed that all populations are reachable or have equal access to computers, smartphones, and the Internet (Deakin and Wakefield Reference Deakin and Wakefield2014; Lobe, Morgan, and Hoffman Reference Lobe, Morgan and Hoffman2020). Relying disproportionately on digital methods automatically disadvantages those studying rural communities, countries without reliable Internet infrastructure, or those with security concerns (or bans) regarding the use of particular platforms (Tanczer et al. Reference Tanczer2020). In the United States, the U.S. Census Bureau estimated that 92% of households had at least one computer, but only 85% had a personal broadband Internet connection and lower income households were more likely to rely on a smartphone for Internet access (Martin Reference Martin2021). By contrast, in South Africa just 38% of the population was estimated to have reliable Internet access in 2021 (Brazil 2022). In many countries, Internet infrastructure has also been leapfrogged by phone providers offering 3G (or higher) Internet access, often resulting in rural areas prioritizing access to smartphones for connectivity (Mothobi and Grzybowski Reference Mothobi and Grzybowski2017). Several studies further show that phone and Internet access in low and middle income countries is gendered; in some countries women are significantly less likely to have mobile access than men, and even less access to mobile data (Reñosa et al. Reference Reñosa2021). Hence, context matters greatly when considering remote fieldwork.
Logistical features must equally be deliberated, such as the requirement of participants’ telephone numbers when using FaceTime or WhatsApp or the need to be “friends”’ on Facebook and “followers” on Instagram to access private content. Moreover, the costs associated with certain platforms and features, such as premium Zoom accounts or Microsoft Teams for calls without time constraints, may prove to be a hindrance for both participants and scholars. As legislation varies between countries regarding which platforms and applications are available and supported by state governments, this too can prove challenging when attempting to recruit and access international participants (King, Pan, and Roberts Reference King, Pan and Roberts2013; Hobbs Reference Hobbs and Roberts2018). While participant recruitment of all types is able to reach certain demographics more easily than others, scholars conducting remote research must be extra cognizant and transparent about these factors when describing their recruitment strategies and considering their populations under study.
In addition to recruitment concerns, remote fieldwork additionally impacts researchers’ rapport with participants, as well as affecting the kinds of answers individuals may give in interviews or surveys. Some participants may fear engaging with technology, in addition to speaking with researchers in a digital setting. For scholars recruiting survey respondents online, the anonymity often provided to respondents behind a username has trade-offs; anonymity may lead respondents to be more open to sharing sensitive information or opinions, but can also increase the number of fraudulent and dishonest respondents (Teitcher et al. Reference Teitcher2015). For researchers conducting interviews through platforms like Skype, Zoom or Microsoft Teams, remote approaches also generate a degree of separation between researchers and their participants, therefore limiting interpersonal connections (Oliffe et al. Reference Oliffe, Kelly, Montaner and Ko2021). Whereas the ability to mute or keep video cameras off may offer additional comfort to those interviewing virtually, conversations may be choppy due to time lags in audio, and pacing consequently challenging to manage (Mwambari, Purdeková, and Bisoka Reference Mwambari, Purdeková and Bisoka2021). Although overcoming the physical distance between researchers and their participants, mediated technologies simultaneously emphasize the fact that the actors are not located in the same place (Howlett Reference Howlett2021). Interviewing participants through telecommunications software like Skype also limits the capacity of researchers to fully observe the context of an interview, making it difficult to view and acknowledge body language or non-verbal cues in conversations (Reñosa et al. Reference Reñosa2021). Collectively, these challenges may both severely impact who we reach, and the quality and depth of data yielded from such approaches.
Perhaps most important, privacy concerns are heightened in all remote fieldwork contexts when researchers cannot control for everything within their surroundings. Children, family members, and co-workers interrupting or overhearing conversations add additional concerns when meetings are held from home or in shared work spaces, as do larger unpredictable disruptions like construction, weather, and power outages. These issues are amplified if computers are shared by family members or children in the same household, as conversations or research materials may be inadvertently or mistakenly accessed or shared publicly. These concerns can certainly impact the conduct of the researcher, but also the willingness of participants to speak candidly. As neither party can assure the other that the conversation will not be overheard or recorded, as would be the case during in-person fieldwork, privacy and data management concerns become particularly concerning for those conducting research on sensitive topics. When conducting interviews of any sort, but especially remotely, it is hence the responsibility of the researcher to protect the content of the conversations and ensure interlocutors’ anonymity however necessary (Jenner and Myers Reference Jenner and Myers2019), but also to reassure participants of the measures being taken to do so.
While remote research indeed comes with its own challenges to recruitment and access, it must nevertheless be stated that there are benefits in using these approaches. In particular, scholars may have better luck using digital methods to access participants who travel often or have busy schedules, such as political elites or business officials, if the meetings can be conducted while participants are between places or taking breaks. The increased use of social media and telecommunications software globally during the COVID-19 pandemic also saw many more individuals and organizations providing livestreams of events and public meetings, meaning that researchers have been able to attend in real-time without the travel and clearances required to be physically present.Footnote 5 Likewise, participatory video technology can serve as an alternative means of interviewing, allowing study participants to record themselves on their smartphones (Marzi Reference Marzi2021). As these techniques are likely to stay now that the infrastructure to facilitate them exists, they have the benefit of being incredibly low-cost and effective means for researchers to both connect with new participants and remain in contact with others long after they “exit” the field, whether physically or from afar.
Beyond interviewing and other forms of participant engagement, remote fieldwork has opened scholars’ access to a wealth of materials, especially for those using archival material. Virtual reading rooms and digitized archives have proven effective for the start and continuation of research projects (MacLean et al. Reference MacLean, Rahman, Turner and Corbett2020). Search engines, web scraping tools, and optical character recognition software have furthermore facilitated the processing of documents much faster than human capacity, while also enabling researchers to recover new perspectives of key historical events (Crossen-White Reference Crossen-White2015). Yet, in spite of the advantages, it is still important for researchers to remain critical and evaluate whether (and to what extent) the digitization of materials is selective, as is often the case in digitized archives, as well as how digitally accessing materials may affect one’s evaluation of them (Voges, Märgner, and Fingscheidt Reference Voges, Märgner and Fingscheidt2009). Many archives have had limited resources to fully digitize their collections, and digitization requests submitted by researchers can (and have) generated a backlog that archives and libraries struggle to process, especially in the wake of pandemic closures (Miller Reference Miller2013). As a result, an assessment of what material may be missing, as well as how one assesses the material, is necessary for ensuring the transparency and replicability of our findings, particularly as our interpretations may change. For instance, scrolling through individual pages on a personal computer through a virtual finding aid does not replicate the experience of physically flipping through a box of materials in an archive and holding documents to analyze them. Digitized documents can also be challenging for some researchers to work with as they can prove taxing on their eyes; this concern may be person-specific but is nonetheless applicable for all forms of digital data collection, and important given the increased use of screens in our work-from-home lives. For these reasons, the unique benefits and limitations to recruitment and access must be recognized when research is conducted remotely.
Informed Consent and Data Security
Utilizing remote and digital methods for fieldwork additionally creates new challenges for informed consent and data security. Informed consent and the anonymization of data is especially pertinent in situations wherein participants speak on certain issues that, if made public, may lead to professional repercussions, personal scandals, or physical harm (Mosley Reference Mosley2013). For research involving humans prior to the COVID-19 pandemic, IRBs accordingly strictly required informed consent and data management plans prioritizing the protection of participants’ identities in line with federal legislation; in the United States, the American Political Science Association (APSA), for example, defers to the 1991 Federal Policy for the Protection of Human Subjects (“Guide to Ethics in Political Science” 1992). Ensuring that research participants are able to fully and independently engage in studies as informed and consenting individuals has hence become both a significant ethical concern for scholars conducting field research, as well as a legal and moral consideration.
Yet with the widespread use of the Internet and modern communications technologies, especially since March 2020, social science researchers have been made increasingly aware of the risks of poor data protection policies; leaked interviews and poorly safeguarded information have been rapidly distributed globally in cases when scholars, and also the general public, have not carefully handled their data (Aldridge, Medina, and Ralphs Reference Aldridge, Medina and Ralphs2010). Some apps have additionally made it easy to geolocate events and individuals by tracing their personal home and IP addresses, risking de-anonymization (van Baalen Reference van Baalen2018), while other communications monitoring software allows state governments to actively track citizens’ behavior and Internet history on their smartphones and personal computers (Lyall Reference Lyall, Bennett and Checkel2015). Governments equally have access to citizens’ public social media accounts, as well as are able to infer data from companies such as Meta or Amazon using machine-learning tools to determine consumer profiles, interests, and types (Grimm et al. Reference Grimm2020). For these reasons, participants’ fears of surveillance, and surveillance practices themselves, can significantly impact the communities under study in political research. Hajer Al-Faham’s Reference Al-Faham2021 study on American Muslims underscores this reality in showing how access to participants can be strained when local community organizations are reluctant to engage with researchers given their fear of local police and the FBI (Al-Faham Reference Al-Faham2021). Although in-person ethnographic work poses its own challenges to participants’ safety and security, digital tools for data collection present researchers with new considerations to ensure the security of their participants.
Still, ethics in the context of digital and remote field research has typically been defined as distinct from in-person data collection because the use of digital technologies to collect data has not regularly been recognized as comparable, this is exemplified by van Baalen (Reference van Baalen2018), who defines field research as “the collection or generation of data or experiences in a physical socio-political or geographical site where the researcher spends time, which excludes online research and remote data collection” (van Baalen Reference van Baalen2018, 2). Following from here, remote interviews, focus groups, and the collection of virtual data from public chat forums and pages, are understood as data collected from the “public” domain—that is, data that are accessible to anyone able to search and download—rather than equivalent to materials gathered in person for research purposes (Sanders Reference Sanders2021). Nevertheless, van Baalen’s definition is not the only accepted understanding of fieldwork, and it predates the COVID-19 pandemic when data in online public domains had yet to be widely recognized as valuable for research purposes. This means that IRB standards may not always apply nor be deemed necessary for these approaches to data collection, and even may be excused if a participant is thought to be acting in a public domain rather than in a research-related capacity.
These oversights have been emphasized in previous works. For example, in a study of Internet censorship in China, researchers assessed whether local networks could access specific websites using TCP/IP side channels on a continuous basis. In effect, the study generated access requests to specific domains when a user in China visited a website with the study’s code embedded in it. In doing so, the study effectively generated internet traffic from an individual to censored websites without them being personally aware. The initial study was proposed, developed, and initiated without IRB approval given that computers, independent of human action, were generating the data used by the researchers. However, the study was eventually subject to the IRB and the scholars were required to develop their program in a way that accessed Internet infrastructure, such as general servers, that were not linked to personal or household computers (Pearce et al. Reference Pearce2018). This discussion underscores the importance of ensuring our participants are fully and independently consenting individuals, even if data access may not require face-to-face interactions.
Nevertheless, there is still no widely recognized best approach to obtaining informed and independent consent in the context of virtual communities, public posts on social media, or private group chats. This raises two concerns: first, whether the content of public posts systematically differs from private ones and, second, whether consent is required to use such content and how it can be obtained. Importantly, informed consent procedures for in-person research do allow researchers to adapt to participants’ needs; for example, verbal consent is regularly accepted when conducting research with illiterate populations or in societies where written notes and signatures may be, or at least feel, unsafe for participants and lead them to decline participation, such as in some post-Soviet contexts (Jacobs et al. Reference Jacobs2021). Although beneficence and interpersonal connections may appear more difficult to establish during digital field research than when we engage directly with research participants in a field site, a false sense of consent by virtue of our participants’ online engagement or posts cannot be accepted. Rather, it is important for scholars using digital methods to think of appropriate measures to similarly adapt to the needs of the populations we study.
Gaining informed consent to record conversations and to attribute quotes in any named sense likewise poses its own difficulties when research is conducted remotely. Auto-transcription services, such those available with business or institutional Zoom accounts, are useful for research in that they can provide scholars with downloadable transcripts immediately after a meeting; however, they break end-to-end secure networks in requiring that meetings be saved as a cloud recording (Chia, Ghavifekr, and Razak Reference Chia, Ghavifekr and Razak2021; Zoom 2022). For many IRBs, linking video recordings to a cloud server beyond the researcher’s control breaches data management plan prerequisites and risks personal data being accessed by external sources. Broader security concerns, including impersonation risks, have also been documented in evaluations of internal white papers drafted by Zoom; these memos noted that impersonation risks exist, which may especially pose concerns for researchers conducting focus groups over any telecommunications platforms (Archibald et al. Reference Archibald, Ambagtsheer, Casey and Lawless2019). For researchers relying on remote fieldwork to conduct interviews on sensitive topics, the best option is therefore not to record any meetings, or to record with a separate device not directly connected to Wi-Fi or the cloud.
As standards around informed consent for digital and remote research, and also the use of data collected through these means, are still very much emerging, scholars must, at minimum, independently acknowledge potential harm to participants. In cases where researchers can differentiate between what an interlocutor observes and their opinion on what they have observed, consent may only be required to use the latter by their IRB. Barratt and Maddox (Reference Barratt and Maddox2016) demonstrate this in showing that careful assessment of data protections and IRB approval can enable scholars to study illicit virtual communities. In their digital ethnography of the Silk Road and associated crypto markets, the authors transparently discussed their identities as researchers with users, choosing to use their real names in order to build rapport (Barratt and Maddox Reference Barratt and Maddox2016, 706-707). Although the authors did not disclose their method for obtaining informed consent in their study, their decision to rely on direct quotes from users’ posts, referencing only the forum and date of the posts, still worked to protect the identity of the users. This approach stands in contrast to Gehl (Reference Gehl2016)’s study of darknet social media, where the researcher used a pseudonym when engaging with participants online and obtained informed consent only before interviews (Gehl Reference Gehl2016). While evaluating the need for, and the best approach for obtaining consent in the context of data collection and dissemination must align with our evolving understandings of harm and risk, best practice should have scholars evaluating how any expressed opinion and personal information in their possession, regardless of how it was collected, guarantees the safety and dignity of the participants and communities involved.
Publishing and Disseminating Findings
The ethics of what we publish and disseminate following remote and digital data collection is also central to questions about our ethical obligations. As scholars, we face immense pressure to publish our data quickly following collection—both for professional incentives and out of fear that the data will be published by others or become outdated. The COVID-19 pandemic has only increased these anxieties, as well as adding others related to data and resources access, in requiring many scholars to work from home, homeschool their children, and adopt increased caring responsibilities; the pressure is only further compounded with lengthier publication processes due to difficulties in finding peer reviewers. While these ever-growing stresses are slightly reduced by the availability of digital technologies for conducting research remotely—the same tools that allow us to mitigate the health risks to our participants—the ethics of what should be published and when must not be ignored in our urgency to publish. Knott (Reference Knott2019) highlights this argument when discussing her work in Crimea, underscoring that scholars must think ethically when disseminating findings about dangerous research environments due to the more obvious physical and material risks to interlocutors. In her piece, she asserts that scholars must ask serious ethical questions about “what should be published [away from the field] to minimize the potential for harm to and to ensure the dignity and respect of research participants” (Knott Reference Knott2019, 146). In following from her call to consider ethics when both collecting and disseminating data connected to dynamic field environments, we can see that serious ethical concerns likewise exist when publishing data collected remotely. Table 1 highlights three stages at which authors have an obligation to the discipline to be transparent about their research design, data collection, and ethics management in a given study.
Table 1 Researcher responsibilities regarding transparency
Note: While we prioritize transparency and ethics in digital and remote fieldwork in this paper, these responsibilities apply to all research, regardless of methodological approach. For those conducting qualitative studies in particular, transparency with regards to research design choices and data generating processes allows readers to better evaluate potential biases and study limitations.
Beyond participant recruitment and informed consent, discussions of transparency and replicability have arisen in many conversations amongst qualitative researchers since 2010, especially those relying on fieldwork projects based on some degree of subjective data collection and interpretation. The Data Access and Research Transparency (DA–RT) initiative and the Journal Editors Transparency Statement (JETS) consequently built on broader calls within the discipline of political science to consider “how data [are] accessed, produced, and analyzed” as “an ethical obligation to the discipline” (Knott Reference Knott2019, 148). While the Qualitative Transparency Deliberations (QTD) report addressed the DA–RT initiative in the context of field research, whether conducted virtually or in person, it was completed in 2018, before the COVID-19 pandemic. Accordingly, a researcher’s access to a field site and recruitment strategy, regardless of whether involving in person or remote techniques, may not be clear-cut or easy for an outsider to understand during peer review. As the use of digital tools has increased faster than the discipline’s ability to outline best practices regarding data access and ethics, reviewers may consequently devalue studies using online public materials or interviews and focus groups in favor of those following traditional “shoe leather” approaches, assuming that virtual methods influence how people self-present themselves (Freedman Reference Freedman1991; Kiesler and Sproull Reference Kiesler and Sproull1992; Spears and Lea Reference Spears and Lea1994). While in-person field research has historically been viewed as “the gold standard against which the performance of computer-mediated interaction is judged,” digital methods have now proven to be equally valid and legitimate approaches to research (Hine Reference Hine2005, 4). Immersion and embeddedness as a standard also do not require a set amount of time in a field site or a particular number of sites evaluated; it can very much be achieved remotely. While digital and remote methods, like all methods, have their limitations, what ultimately matters in the conduct of ethical research is a demonstration of the depth and quality of the data collected, and likewise, the thoroughness of evaluation and interpretation of results offered by a scholar. The onus is thus on the researcher to demonstrate how the data they collected answer the research question being asked, regardless of the method used, and also on the profession to ensure data are evaluated fairly.
At this time, the transparency standards in political science routinely requires researchers to discuss how their data—whether qualitative or quantitative—were collected, what was beyond reach, why any gaps may impact the results, and how weaknesses in the project allow for future analyses. Conducting transparent digital and remote field research accordingly requires researchers to disclose their personal biases and any interventions that may lead them away from strict objectivity. This reality is no different than the disclosure of funding sources that may generate biases, or how ethnographers disclose when they personally intervened in a study. For scholars relying on digital archives and virtual interviews, it is equally important to evaluate what is missing from a database, what subset of a population is participating, and how the inclusion and exclusion of certain groups can impact their results. Doing so undoubtedly requires a critical assessment of what may be lacking from a study, which is equivalent to researchers’ evaluations of what documents are missing from a folio in an archive or whom may be absent or overlooked from an ethnography. When observing or interacting with interlocutors in online spaces, just like face-to-face interviews, focus groups, and participant observations, it is equally as important to realize who does and does not engage, as it reveals various hierarchies, relationships, and inclusion/exclusion dynamics. In both in-person and online approaches to data collection, then, it remains fundamental for scholars to consider who may be missing or silenced by others, and by our methodological decisions.
In addition to transparency, replicability has been increasingly held as a key standard of social science scholarship (King, Keohane, and Verba Reference King, Keohane and Verba1995). With efforts by the political science discipline to emulate the natural sciences, scholars relying on materials collected in the field have often been asked to consider how best to render their work replicable while ensuring anonymity and minimal risk to those involved. The final report of QTD Working Group III.2 discussing ethnographic and participant observation methods argued that access to field notes, especially chat or interview transcripts, may harm research subjects especially if research is conducted in non-public digital communities like WeChat and WhatsApp or via mediated but private interviews, such as on Zoom, Skype, and FaceTime (Jacobs et al. Reference Jacobs2021). Although standard issues of access to field notes apply to other methods for data collection, they may be blurred when involving public social media like Twitter, Reddit, and Facebook accounts, or where all or partial aspects of respondent data can be scraped or provided by the account’s host. While existing methods guides have prioritized transparency with an eye to replicability, we argue that transparency when using remote and digital methods matters more in terms of process, as replicability and transparency must ultimately come second to the potential harms posed to the studied participants and their communities.
Much like the ethical issues of access and informed consent, then, the expanded consideration of “form, degree, and loci” of potential harm must also be recognized in the context of remote research. In addition to upholding standards of privacy, anonymity, and confidentiality to minimize risks, scholars must evaluate both what is safe and what is feasible. Ensuring that these ethical obligations are upheld may indeed require field researchers to assess the trade-offs of their approaches, defend what is feasible given the conditions they are working within, and actively consider the physical and emotional risks posed to them and their participants. In some instances, this may involve altering or even canceling aspects of the data collection, if it means minimizing potential harms to participants. As the QTD Working report noted, field researchers may not be required to provide a full “replication file” of their data if they are sensitive and potential harm to interlocutors is high. Because scholars often lose control of their material once it is published, in terms of its interpretation, use, and impact, additional care must accordingly be taken to consider the sensitivity of the data, and the ways it may be harmful for the people and places we study (Fujii Reference Fujii2012). In sum, the new methodological approaches to data collection, while advantageous in many ways, fundamentally challenge the ways we understand ethics both procedurally and in practice.
Conclusion: “Returning” to the Field in New Ways
Since March 2020, the use of remote fieldwork and digital methods for research has increased exponentially. While indeed in response to the biological risks posed by COVID-19, and the associated barriers to travel and access to people and resources, the growing use of these methods is likewise connected to our expanded understandings of harm. Whereas harm was previously understood as the psychological risks of research on ethics, it has grown to include the wider health and wellbeing of the people and societies we study. Whereas the methodological techniques now used to conduct remote data collection pre-date the pandemic, their broader use and recent widespread acceptance in political science has opened access to new data, and allowed scholars to conduct a wealth of new research that would not have otherwise been feasible, especially for those with limited resources or little capacity to travel to field sites for extended periods of time. Rather than abandoning certain research topics, scholars have thus been able to use telecommunications platforms such as Zoom, Skype, or Microsoft Teams to facilitate interviews or focus groups; utilize digital repositories and online archives rather than physically accessing archives; and analyze personal or anonymous accounts on various social media platforms for participant observations. Looking ahead to a post-pandemic world, researchers will undoubtedly continue their use of remote fieldwork and digital data collection techniques as they have proven to be useful and distinct methods in their own right.
But while these digital tools have proven very effective for political science research, conversations about ethics within the discipline have not yet demonstrated “ongoing ethical reflection and growth” or even overtly acknowledged our ethical responsibilities in light of these new methodological approaches (Hine Reference Hine2005, 4). Although the adoption of remote approaches has led many scholars to believe that ethical concerns relating to harm are mitigated, especially those connected to disease transmission, these methods carry with them their own set of ethical challenges, particularly with regard to participant recruitment, informed consent, and publication. For instance, remote interviews and focus groups facilitate access to certain populations, but limit engagement with others, particularly among female, low income, rural, or senior communities. Cellular technologies and social media enable researchers to observe and stay in contact with their participants from afar, yet such platforms still present privacy and security concerns if surveilled by governments or accessed by software companies. Digital repositories and archives are readily accessible to scholars with reliable Internet and institutional access, but digitization has not uniformly included all sources. Defining what are “public” and “private” domains, and ensuring our interlocutors are fully aware and consenting participants, is even more important now that we cannot be located together in the same space. Remote methods will accordingly always limit the expense of travel and exposure risk to trauma or disease; however, they may pose additional security concerns and limit deeper engagement. As political science increasingly relies on these methods, researchers have an ethical responsibility to the discipline to defend their approaches, note their limitations, and transparently outline the data-generating processes for future scholars (Fox et al. Reference Fox2021, 776).
Even so, it must be recognized that while we may be aware of our ethical responsibilities once we have “exited” the field or simply conducted remotely away from it, “returning” may not ever again be possible. Aside from the potential health, emotional, and physical risks to either or ourselves and our participants, entering the field in a post-pandemic world may not be feasible, or at all possible, for graduate students and early-career researchers with limited time and financial resources to conduct their research. Likewise, many academics relying on grants with set timelines for deliverables will not always able to visit, return to, or even leave the field in the case of local virus or disease outbreaks, regardless of whether these are in their home communities or the field. Finally, and as Knott (Reference Knott2019) reminds us, dynamic events fundamentally alter our field sites so that our “return” will not be to the same field that we “exited” from. In some cases, volatile situations may disrupt or change our fields’ physicality, such as regime changes, economic growth or decline, environmental disasters, and war. At the same time, our new digital “fields” are equally dynamic, even while not necessarily explicitly considered. For instance, chats, Tweets, and posts can be deleted, websites may be removed, and server issues can prevent the archiving and retrieval of web-based materials. For these reasons, we must more actively “imagine, or predict, what changes may or may not occur after we have left the field,” however we define our “fields,” as dynamic contexts are intrinsic to political science research (MacLean et al. Reference MacLean, Rahman, Turner and Corbett2020, 5-6).
Since research ethics have predominantly been understood in the context of researching sensitive topics, working with marginalized/vulnerable populations, or situating ourselves in dynamic or dangerous environments, this paper has started a broader conversation in the discipline of political science on the ethics of remote fieldwork and digital data. While traditional definitions of fieldwork have prioritized data collection away from one’s home institution, remote fieldwork approaches blur what has been traditionally defined as a rigid boundary between “home” and the “field.” Although not explored here, the ontological implications of this merger warrant further analysis. Similarly, and as our conceptions of harm have expanded, the researcher is also included in these developments; studies show that field research increases the likelihood of researchers being directly exposed to traumatic incidents, secondary trauma, or compassion fatigue due to engagement with participant traumas via interviews or source material (Knott Reference Knott2019, 140). This isolation and other mental health concerns are not specific to field research in dangerous contexts and are just as prescient for those conducting remote fieldwork (Hummel and Kurd Reference Hummel and El Kurd2021). Understanding how best to limit potential harm to participants and researchers, including those relating to mental health, warrants much deeper attention. Leaning on insights from parallel social sciences such as sociology and the subfield of human-computer interaction may yield further insights into how political science can best approach both the ethics of remote fieldwork and the challenges of conducting qualitative research from home. While it remains the responsibility of researchers to consider the ethical questions of their research projects, it is likewise necessary for the larger discipline to recognize and uphold consistent ethical standards to mitigate against new and ongoing risks to health. As such, a much greater emphasis must be placed not only on ethics “in practice” and “beyond the field,” but also on research conducted digitally and remotely.
Acknowledgments
The authors thank our six anonymous reviewers and the editorial team at Perspectives on Politics for their constructive comments and feedback. They also extend special thanks to Dr. Leanne C. Powner and Professor Layna Mosley, for their useful comments and suggestions on the first draft of this paper. Additional thanks are due to participants from Princeton University’s Qualitative Research Colloquium for their insight on the ideas behind this project.
Open access
