Algorithmic Hauntings: miscarriage and grief in online social media

Lead Research Organisation: University of York
Department Name: Sociology

Abstract

A majority of prospective parents in 'advanced economies' now use online social media to enhance their ante-natal experience. These media use artificial intelligence (AI) to analyse users' preferences, and recommend to them the most "relevant" content. Such AI can be beneficial in pregnancy: a short-cut to maternity products and services that meet individual needs. However, when miscarriage occurs this AI is slow to adapt; adverts for baby clothes or notifications from 'pregnancy apps' can 'haunt' the bereaved online for months after loss. These "hauntings" complicate grief, and compromise the safety of vulnerable people online. They are likely widespread: social media use is ubiquitous, while 1 in 4 UK women miscarry in the first trimester, with black and minority ethnic women disproportionately affected.

This project will investigate the experience of women "haunted" by online pregnancy content post-miscarriage through three key research questions:

How are women "haunted" by online pregnancy content, post-miscarriage?
What causes these "hauntings", and what social media regulation exists or is needed?
What role does ethnicity play in how these "hauntings" are experienced?

Summary

Studies indicate that 1 in 3 women develop PTSD following miscarriage; that BAME women are 40% more likely to miscarry, and more likely to develop severe depression post-miscarriage. In this context, the "haunting" of the bereaved through online pregnancy content is a significant ordeal; further complicating grief; exacerbating existing mental ill-health. Despite this, these "hauntings" remain largely unexplored in sociological research.

It is unclear what protections exist for those "haunted" by pregnancy content online. While a new Online Safety Bill aims to limit illegal material in social media, 'legal but harmful' content of the kind faced by women post-miscarriage will remain unregulated. Social media users can control what they see online through in-built settings, but these are hard to locate, and typically adjusted after 'online harm' has occurred. At the same time, 'pregnancy apps' are increasingly prescribed by NHS midwives, with little guidance on their safe use. The impact of these tools post-miscarriage is likely complex, but has yet to be explored.

This research will examine the experience of women "haunted" by online pregnancy content post-miscarriage; how women comprehend these "hauntings"; and what role they play in the construction of grief and remembrance. It will examine the AI from which these "hauntings" arise, and ask what protections for women "haunted" by pregnancy content exist in social media policy and regulation. It will explore the use of social media in midwifery; asking whether such practice accounts for its more problematic effects. Lastly, it will consider how ethnicity informs the experience of those "haunted", and how inequalities in this area can be addressed.

This research builds on previous Postgraduate research on the use of Facebook in the memorialisation of the dead. It will be supported through key research centres at the University of York: Death and Culture Network, Grief: A Study of Human Emotional Experience, and ALGorithms - Algorithms, Grief and Loss. It will benefit from existing links between these networks and social media companies, and the pregnancy loss charity Tommy's. This research offers a unique contribution to research in the sociology of death, digital media studies, and mental health. It will impact future regulation of social media, and organisations campaigning for social media equity, such as Glitch, and Mind. The project will deliver an accessible training program on 'social media and miscarriage', designed for midwifery and mental health professionals. This training will pinpoint risks inherent to social media in the context of miscarriage, and how their use in health settings can be improved.

People

ORCID iD

Paul Ord (Student)

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
ES/P000746/1 01/10/2017 30/09/2027
2890103 Studentship ES/P000746/1 01/10/2023 30/09/2027 Paul Ord