Reranking partisan animosity in algorithmic social media feeds alters affective polarization

2024/11/22 by Tiziano Piccardi, Martin Saveski, Piccardi, Tiziano +10 · 15 voices · 11 citations
Computer Science · Social Sciences · #Sentiment Analysis and Opinion Mining #Social Media and Politics #Computational and Text Analysis Methods

paper · pdf · doi:10.1126/science.adu5584

Abstract

Today, social media platforms hold the sole power to study the effects of feed-ranking algorithms. We developed a platform-independent method that reranks participants' feeds in real time and used this method to conduct a preregistered 10-day field experiment with 1256 participants on X during the 2024 US presidential campaign. Our experiment used a large language model to rerank posts that expressed antidemocratic attitudes and partisan animosity (AAPA). Decreasing or increasing AAPA exposure shifted out-party partisan animosity by more than 2 points on a 100-point feeling thermometer, with no detectable differences across party lines, providing causal evidence that exposure to AAPA content alters affective polarization. This work establishes a method to study feed algorithms without requiring platform cooperation, enabling independent evaluation of ranking interventions in naturalistic settings.

Cited by

• Cooperative Evolutionary Pressure and Diminishing Returns Might Explain the Fermi Paradox: On What Super-AIs Are Like
• Political attitudes differ but share a common low-dimensional structure across social media and survey data
• Contextualizing Misinformation: A User-Centric Approach to Linguistic and Topical Patterns in News Consumption
• Moving towards informative and actionable social media research
• Platform architecture determines whether recommendation algorithms can shape information quality on social media
• "F*** You Biden": Cross-Partisan Electoral Toxicity on X
• The political effects of X’s feed algorithm
• Ad-based social media interventions increase belief accuracy and generate pro-social opinions among non-news readers
• CTRL-Rec: Controlling Recommender Systems With Natural Language
• Social technologies need societal alignment
• Political polarization in low- and middle-income countries
• Identifying Constructive Conflict in Online Discussions through Controversial yet Toxicity Resilient Posts

Discussions

• Our experiment shows it’s possible to reduce political hostility by designing feed algorithms that account for societal impact. Takeaway: Social media platforms have the power to foster healthier dem [bsky, 180 points, 11 comments]
• When X feeds experimentally show people fewer posts with antidemocratic attitudes, people feel more positively toward outpartisans, finds @tiziano.bsky.social @msaveski.bsky.social @jiachenyan.bsky.s [bsky, 26 points, 1 comments]
• Esto demuestra que las empresas de tecnología realmente pueden tener una fuerte influencia en cómo percibimos a los demás. arxiv.org/pdf/2411.14652 [bsky, 5 points, 1 comments]
• Joint work with @tiziano.bsky.social, @jiachenyan.bsky.social, Jeff Hancock, Jeanne Tsai, @mbernst.bsky.social Link: doi.org/10.1126/scie... And a very thoughtful perspective by @jennyallen.bsky.so [bsky, 4 points, 2 comments]
• Given the recent evidence that Twitter does cook brains and create bias, the people finding no effects should be presumed to be honest arxiv.org/abs/2411.14652 [bsky, 4 points, 1 comments]
• An interesting way to reduce affective polarization! doi.org/10.1126/scie... [bsky, 3 points, 0 comments]
• 翻訳記事しか見てないけど、これなかなかショッキングな報告なのでは… あとで原著論文読もう（汗） arxiv.org/abs/2411.14652 [bsky, 3 points, 0 comments]
• Can social media platforms foster healthier democratic discourse? Scientists ran an experiment, adjusting exposure to polarizing posts. Result: Political hostility can be reduced by redesigning feed [bsky, 1 points, 0 comments]
• 🚨New study on X/Twitter with 1,256 consented participants concludes: "Exposure to AAPA content also results in an immediate increase in negative emotions, such as sadness and anger" arxiv.org/abs/241 [bsky, 1 points, 1 comments]
• Social Media Algorithms Can Shape Affective Polarization via Exposure to Antidemocratic Attitudes and Partisan Animosity Tiziano Piccardi, Martin Saveski, Chenyan Jia, Jeffrey T. Hancock, Jeanne L. Ts [bsky, 1 points, 0 comments]
• Social Media Algorithms Can Shape Affective Polarization via Exposure to Antidemocratic Attitudes and Partisan Animosity arxiv.org/abs/2411.14652 [bsky, 0 points, 0 comments]
• arxiv.org/abs/2411.14652 [bsky, 0 points, 0 comments]
• プレプリント Source and Image Credits: Piccardi, Tiziano, et al. “Social Media Algorithms Can Shape Affective Polarization via Exposure to Antidemocratic Attitudes and Partisan Animosity.” arXiv preprint ar [bsky, 0 points, 0 comments]
• Let's see if this gets momentum and social networks realise they can make money without turning people against each other. arxiv.org/abs/2411.14652 [bsky, 0 points, 0 comments]
• X could manipulate its users. arxiv.org/abs/2411.14652 [bsky, 0 points, 0 comments]

Related

• How do social media feed algorithms affect attitudes and behavior in an election campaign?
• The political effects of X’s feed algorithm
• Engagement, User Satisfaction, and the Amplification of Divisive Content on Social Media
• Quantifying Political Partisanship for Cross-Platform Analyses
• Divided by the Algorithm? The (Limited) Effects of Content- and Sentiment-Based News Recommendation on Affective, Ideological, and Perceived Polarization