Beyond Social Support: Spiritual Support as a Novel Design Dimension in Sociotechnical Systems

Abstract

For people facing life-threatening physical and/or mental illness, matters of spirituality or religion often assume a role of elevated importance in their lives. However, spirituality is an understudied topic in Social Computing, and has largely been omitted or minimized in research narratives and trajectories related to online health. Taking a human-centered design approach that seeks to honor people's core values and beliefs, this dissertation contributes an empirically-derived theoretical perspective that views spirituality as a crucial underlying dimension in sociotechnical systems for online health and social support, such as Online Health Communities (OHCs). Because of their widespread adoption and ability to provide users with social support, OHCs are a topic of prominent interest in Human-Computer Interaction and Social Computing. OHCs have been studied across a variety of online spaces, ranging from disease-specific niches on larger social media platforms, to specialized platforms designed specifically for patients and caregivers--such as CaringBridge, a Minnesota-based nonprofit OHC that served over 300,000 people daily in 2019, including 40 million unique users from 237 countries. Two studies in this dissertation were completed in collaboration with CaringBridge, while the third presents Flip*Doubt, a novel prototypical system for crowd-powered cognitive reappraisal. Whereas prior work in OHCs has focused on conventional support categories, the first study provides a content analysis and survey of CaringBridge users that, together, quantitatively distinguish ``prayer support'' as an independent category of crucial importance to users. Because prayer indicates a deeper set of beliefs/values that cannot be captured quantitatively, the second study involves qualitative focus groups with CaringBridge stakeholders. This study contributes a definition of ``spiritual support,'' along with its design implications in OHCs. In particular, one implication is that designers should consider technical mechanisms to provide users with assistance with supportive communication. Thus, the final study of this dissertation, a field deployment of Flip*Doubt, can inform the development of future AI/ML-based systems in OHCs for mental health. All together, these contributions help to shift the lens through which we view social support online, and to guide future work towards creating systems that serve the deepest needs of users.