How emotions spread online following celebrity suicide news
Analyses of resharing behavior reveal distinct temporal patterns in the online propagation of different emotions following tragedies
PLOS
image:
Analyses of resharing behavior reveal distinct temporal patterns in the online propagation of different emotions following tragedies.
view moreCredit: cottonbro studio, Pexels, CC0 (https://creativecommons.org/publicdomain/zero/1.0/)
Reshare cascades on X (formerly Twitter) show how different expressed emotions unfold in the aftermath of celebrity suicides, according to a study published December 10, 2025, in the open-access journal PLOS One by Ehsan Nouri of the University of Virginia, U.S., and colleagues.
Online social media enables the widespread dissemination of user-generated content on an unprecedented scale. It can act as a powerful channel for emotional transmission, and exposure to negative emotions on social media has been shown to adversely affect mental well-being. Research also indicates that negative discourse tends to grow to dominate social media platforms, particularly after shocks and crises.
It's unclear exactly how negative social media content shapes users’ feelings, opinions, and behavior, since a resharing decision does not necessarily indicate that individuals have adopted the underlying emotion. Nonetheless, patterns of resharing over time can reveal how different emotions circulate through an online network and can expose large audiences to forms of negativity such as fear or anger. To examine these circulation patterns, Nouri and colleagues used a language model to extract emotions from more than one million tweets in the aftermath of four celebrity suicides that occurred in 2012 or 2013.
The results showed that 40% of reshare cascades signaled fear as their dominant emotion, followed by 33% as sadness. 11% were neutral, and the remaining 12% indicated either surprise, anger, joy, or disgust as their dominant emotion.
Disgust appeared as the most contagious emotion, associated with rapid, widespread, and persistent patterns of resharing, while fear, despite its arousal, was linked to relatively weak diffusion. Anger and surprise were associated with fast but short-lived clusters of emotion cascades. Joy, while infrequent, endured longer than neutral and negative content when present. Overall, the findings underscore the various emotions expressed after tragic events and how they propagate in distinct ways.
The authors’ analysis is limited by focusing solely on messages reshared on a single platform over a specific time period, and relied on the content of messages to infer their emotions rather than emotions being reported directly. Future research linking social media data with population-level suicide statistics might determine if online emotion propagation could contribute to self-harming copycat behavior.
In the meantime, the study could offer insight into risks linked to emotional amplification in digital environments. For instance, as anger tends to spread rapidly, suicide prevention organizations might seek to act quickly when such content appears online. Sadness and fear, while less “viral”, still expose vulnerable populations to prolonged negativity, while rare positive emotions might offer opportunities to amplify resilience-focused messaging. Together, the results emphasize the need for emotion-specific approaches to both theoretical modeling of and practical intervention in online environments.
The authors add: “In the aftermath of public shocks and tragedies, negative emotions spread widely across social media, and our study shows that different emotions follow distinct pathways in how quickly they spread, how long they keep circulating, and how large they grow. Because this emotional exposure can be harmful for vulnerable users, especially in cases like suicide news, we show how digital platforms may unintentionally amplify risk during sensitive events.”
In your coverage, please use this URL to provide access to the freely available article in PLOS One: https://plos.io/49Ls0C2
Citation: Nouri E, Saraf N, Goh JM, Dasgupta S, Cyr D (2025) Online propagation of emotions: A study of resharing dynamics on social media following celebrity suicides. PLoS One 20(12): e0336134. https://doi.org/10.1371/journal.pone.0336134
Author countries: U.S., Canada
Funding: This study was supported by the Insight Grant from the Social Sciences and Humanities Research Council of Canada Grant Number 435-2013-0641. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Journal
PLOS One
Method of Research
Observational study
Subject of Research
People
Article Title
Online propagation of emotions: A study of resharing dynamics on social media following celebrity suicides
Article Publication Date
10-Dec-2025
No comments:
Post a Comment