Causal inference in ethnographic research: Refining explanations with abductive logic, strength of evidence assessments, and graphical models.

In their classic accounts, anthropological ethnographers developed causal arguments for how specific sociocultural structures and processes shaped human thought, behavior, and experience in particular settings. Despite this history, many contemporary ethnographers avoid establishing in their work direct causal relationships between key variables in the way that, for example, quantitative research relying on experimental or longitudinal data might. As a result, ethnographers in anthropology and other fields have not advanced understandings of how to derive causal explanations from their data, which contrasts with a vibrant "causal revolution" unfolding in the broader social and behavioral sciences. Given this gap in understanding, we aim in the current article to clarify the potential ethnography has for illuminating causal processes related to the cultural influence on human knowledge and practice. We do so by drawing on our ongoing mixed methods ethnographic study of games, play, and avatar identities. In our ethnographic illustrations, we clarify points often left unsaid in both classic anthropological ethnographies and in more contemporary interdisciplinary theorizing on qualitative research methodologies. More specifically, we argue that for ethnographic studies to illuminate causal processes, it is helpful, first, to state the implicit strengths and logic of ethnography and, second, to connect ethnographic practice more fully to now well-developed interdisciplinary approaches to causal inference. In relation to the first point, we highlight the abductive inferential logic of ethnography. Regarding the second point, we connect the ethnographic logic of abduction to what Judea Pearl has called the ladder of causality, where moving from association to intervention to what he calls counterfactual reasoning produces stronger evidence for causal processes. Further, we show how graphical modeling approaches to causal explanation can help ethnographers clarify their thinking. Overall, we offer an alternative vision of ethnography, which contrasts, but nevertheless remains consistent with, currently more dominant interpretive approaches.

Social connection and gene regulation during the COVID-19 pandemic: Divergent patterns for online and in-person interaction.

BACKGROUND: Social connection has been linked to reduced disease risk and enhanced antiviral immunity, but it is unclear whether online social connections have similar effects to those previously documented for in-person/offline social relationships, or whether online connections can substitute for in-person social relations when the latter are restricted. We examined this question in the context of the COVID-19 pandemic, focusing specifically on an immune system gene regulation profile known as the conserved transcriptional response to adversity (CTRA), which is characterized by up-regulation of proinflammatory genes and down-regulation of genes linked to innate antiviral responses and antibody production. METHODS: We analyzed CTRA RNA profiles in blood samples from 142 healthy young adults (69% female, 87% white) during the "social distancing" period of the COVID-19 pandemic. Mixed effect linear models quantified the relation of CTRA gene expression to measures of in-person social connection (number of friends, social eudaimonia, loneliness) and online psychosocial connection (online loneliness, perceived social value in online leisure and educational contexts, and internet use) while controlling for demographic and health factors. RESULTS: Multiple indicators of in-person and generalized social connection were associated with lower CTRA gene expression, whereas no measure of online social connection showed any significant association with CTRA gene expression. CONCLUSION: Experiences of in-person social connection are associated with reduced CTRA gene expression during a period of restricted social interaction. In contrast, online social relationships show no such association. Digitally mediated social relations do not appear to substantially offset the absence of in-person/offline social connection in the context of immune cell gene regulation.