In vivo conceptualisation through in situ communication in virtual ethnographies
During the Covid-19 pandemic, people were required to work from home where feasible, resulting in a shift of everyday work practices such as participating in meetings, attending events, and after-work socialising from the physical to the virtual space. This shift has implications for ethnographic research. In this blog, we share our insights from a virtual team ethnography that not only happened during the pandemic but also focused on exploring practices resulting from the pandemic.
We studied insurance-industry initiatives aimed at developing solutions for sharing the risk of business interruption linked to a pandemic. During our study we became embedded in virtual inter-organisational working groups created to develop these solutions. We participated in four parallel working groups, which convened, according to their own schedules, weekly, fortnightly or monthly. In each of these working groups we participated as members in the work of these groups. We also attended relevant virtual public events and interviewed various stakeholders including insurance firms, industry associations, businesses, and government departments. Prior to engaging in virtual ethnographic research, we already used instant messaging chat groups (e.g., WhatsApp and Signal) on a regular basis to coordinate various projects, actively discuss research, and share social information.
As the pandemic not only affected the organizations we studied but also our own work practices, we developed a habit of updating each other via the chat group on our experience after attending a meeting, an interview, or an event. We shared interesting (as subjectively perceived by the observer) developments, thoughts, and impressions; often prompting responses and further discussions across the team.
This evolved into an increased use of these chat groups for in situ communication, especially when more than one researcher was present at an event or an interview. This enabled us to exchange and also sense-check in real-time our personal reflections and interpretations of being in the field, what would be difficult, if not impossible, in a face-to-face ethnography. For instance, we could share instantly when one of us noticed participants using specific social cues—such as a grimace, raised voice or smile—to express their feelings, or when a particular practice—such as referring problems to some external actor—was repeatedly and purposefully used by some participants. Using the chat enabled us to check if others were observing similar things, giving us in-field ‘memos’ that can be easily retrieved during team discussions or data analysis. Whenever possible we used the chat apps on our laptops, not just our phones, making it easier to simultaneously follow the respective event and the group chat. Still, navigating both is tricky and required practice.
We found such instant communication helpful for in vivo conceptualising in the field. For example, it helped us to develop the concept of ‘insurance schema’ that seemed prevalent across the initiatives we studied. Looking back at our in vivo chat memos, the insurance schema emerged in three steps. First, during data collection and preliminary data analysis, we regularly shared messages of how surprised we were to see insurers not considering actual customers in the solutions they were proposing. For example, during one meeting, we exchanged this chat “Yeah, I think they have some expectations there … [participant] actually said that the list of customer needs is already partially there from the workshops that we had ... To your point: how do they construct the customer (in their insurancy minds).”
We began to understand that insurers held assumptions about customers, which they all seemed to share, rather than question whether these assumptions reflected customers’ actual experiences of the pandemic. These empirical insights prompted step two where we began to zoom in on the specific assumptions during our data collection. For example, we noticed that one underlying insurer assumption was that customers are too price sensitive. In a third step, we were able to abstract these observations and start conceptualising. We interpreted these assumptions as a shared ‘insurance schema’, that acted as a resource in shaping the possible actions that our participants considered. This led us to collect more data but also purposefully search existing data for evidence to verify (or reject) and then define and refine the emerging schema. Below is a screenshot of one of our chats, made during a participant observation attended by three of us, that demonstrates how the insurance schema concept evolved between us:
While chat groups provided opportunities, they also presented some challenges. First, in situ communication can undermine data collection, as important nuances may be lost while chatting with others. We overcame this challenge by ensuring that more than one team member was present for each observation. We also noted that in situ communication did not take place at the very early stages of data collection, as the new field demanded our full attention. Second, in vivo conceptualisation can enable ethnographers to theorise too early, which may bind them to a particular theory. To overcome this challenge, we regularly examined our data through other theories and concepts. We also supplemented our observations and interviews with documentary data to provoke our thinking.
We believe that group chats during fieldwork helped us better communicate as a team, remain engaged with the field, and also conceptualise more effectively. As such, ethnographers may need to embrace virtual modes of conducting ethnographic research, particularly as some form of remote working appears here to stay.
This research project was funded by the Economic and Social Research Council (Grant Ref: ES/V009389/1).