Online propagation of emotions: A study of resharing dynamics on social media following celebrity suicides
Bibliographic record
Abstract
Emotional contagion on social media, particularly following shocking and tragic events, often unfolds through widespread resharing, amplifying affective responses that are typically intense and negative. This study focuses on the context of celebrity suicides, which have the potential to trigger emotional contagion and lead to adverse behavioral outcomes, such as copycat suicides. Using an exhaustive Twitter dataset covering four celebrity suicides, we theorize the propagation of emotional content through a valence-arousal framework, distinguishing emotions based on affective valence (positive or negative) and physiological arousal (high or low). We analyze how distinct emotions embedded in tweets propagate through retweet cascades, treating each tweet and its retweets as a single cascade. Propagation is measured across four cascade dimensions: size, lifetime, speed, and burstiness. Emotions are extracted from over a million tweets and retweets using a BERT-based language model and are used as predictors in regression analyses of the propagation metrics. Our results show that emotional messages propagate in distinct ways after tragic events. Disgust emerges as the most contagious emotion, spreading quickly, widely, and with longevity, while fear, despite its arousal, spreads weakly. Anger and surprise generate fast but short-lived cascades marked by high burstiness. Joy, though less frequent, endures longer than neutral and negative content, reflecting resilience but with lower burstiness. These findings advance research on online emotional propagation by demonstrating that discrete emotions differ significantly even within the same valence-arousal characteristic. They also offer insights for public health strategies to mitigate risks linked to emotional amplification in digital environments.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".