PsyPost
  • Mental Health
  • Social Psychology
  • Cognitive Science
  • Neuroscience
  • About
No Result
View All Result
Join
My Account
PsyPost
No Result
View All Result
Home Exclusive Social Psychology Social Media

New research uncovers a seemingly universal preference for lower-quality news on social media

by Eric W. Dolan
December 23, 2025
Reading Time: 5 mins read
[Adobe Stock]

[Adobe Stock]

Share on TwitterShare on Facebook

A new analysis of millions of social media posts across seven different platforms reveals that the relationship between political content and user engagement is highly dependent on the specific digital environment. The findings suggest that while users tend to engage more with news that aligns with the dominant political orientation of a specific platform, there appears to be a consistent pattern regarding the quality of information.

Across all examined sites, users tended to engage more with lower-quality news sources compared to high-quality sources shared by the same individual. The study, which highlights the fragmented nature of the modern online landscape, was published in the Proceedings of the National Academy of Sciences.

The motivation for this research stems from a need to update the scientific understanding of social media dynamics. For many years, academic inquiry into online behavior relied heavily on data derived from a single platform, most notably Twitter (now X).

This concentration occurred largely because Twitter provided an application programming interface that made data collection relatively accessible for scholars. As a result, many assumptions about how misinformation spreads or how political biases function were based on a potentially unrepresentative sample of the internet. The research team sought to correct this by broadening the scope of analysis to include a diverse array of newer and alternative platforms.

The study was conducted by a collaborative group of researchers from several institutions. The team included Mohsen Mosleh from the University of Oxford and the Massachusetts Institute of Technology, Jennifer Allen from New York University, and David G. Rand from the Massachusetts Institute of Technology and Cornell University.

Their goal was to determine if phenomena such as the “right-wing advantage” in engagement or the rapid spread of falsehoods were universal truths or artifacts of specific platform architectures. They also aimed to understand whether the rise of alternative social media sites has led to the creation of “echo platforms,” where entire user bases segregate themselves by political ideology.

To achieve this, the researchers collected data during January 2024. They focused on seven platforms that allow for the public sharing of news links: X, BlueSky, Mastodon, LinkedIn, TruthSocial, Gab, and GETTR. This selection represents a mix of mainstream networks, professional networking sites, decentralized platforms, and sites that explicitly cater to specific political demographics.

The final dataset included nearly 11 million posts that contained links to external news domains. This large sample provided a comprehensive cross-section of online sharing behaviors.

Google News Preferences Add PsyPost to your preferred sources

The researchers employed a rigorous set of measures to evaluate the content within these posts. To assess the quality of the news being shared, they did not rely on their own subjective judgments. Instead, they utilized a set of reliability ratings for 11,520 news domains. These ratings were generated through a “wisdom of crowds” methodology that aggregated evaluations from professional fact-checkers, journalists, and academics. This system allowed the team to assign a quality score to the publisher of each link, serving as a proxy for the likely accuracy of the content.

In addition to quality, the team needed to quantify the political leaning of the news sources. They utilized a sophisticated large language model to estimate the political alignment of each domain. The model was asked to rate domains on a scale ranging from strongly liberal to strongly conservative.

To ensure the validity of these AI-generated estimates, the researchers cross-referenced them with established political benchmarks and found a high degree of correlation. This allowed them to categorize content as left-leaning, right-leaning, or neutral with a high degree of confidence.

The primary statistical method used in the study was a linear regression analysis that incorporated user fixed effects. This is a statistical technique designed to control for variables that remain constant for each individual. By comparing a user’s posts only against other posts by the same user, the researchers effectively removed the influence of popularity. It did not matter if a user had ten followers or ten million. The study measured whether a specific user received more engagement than usual when they shared a specific type of content.

The results regarding political polarization challenged the idea of a universal advantage for conservative content. The data indicated that the political lean of the most engaging content generally matched the political lean of the platform’s user base.

On platforms known for attracting conservative users, such as TruthSocial, Gab, and GETTR, right-leaning news sources garnered significantly more engagement. On platforms with more liberal or neutral populations, such as BlueSky, Mastodon, and LinkedIn, left-leaning news attracted more likes and shares.

This finding supports the hypothesis of “echo platforms.” In the past, researchers worried about echo chambers forming within a single site like Facebook. The current landscape suggests a migration where users choose entire platforms that align with their views.

The researchers found a strong correlation between the average political lean of a platform and the type of content that gets rewarded with engagement. This implies that the “right-wing advantage” observed in earlier studies of Twitter and Facebook may have been a product of those specific user bases rather than an inherent property of social media.

While political engagement varied by platform, the findings regarding news quality were remarkably consistent. The researchers discovered that on all seven platforms, posts containing links to lower-quality news domains received more engagement than posts linking to high-quality domains. This pattern held true regardless of whether the platform was considered left-leaning, right-leaning, or neutral. It was observed on sites with complex algorithmic feeds as well as on Mastodon, which displays posts in chronological order.

The magnitude of this effect was notable. The analysis showed that a user’s posts linking to the lowest-quality sites received approximately seven percent more engagement than their posts linking to high-quality sites. This effect was robust even when controlling for the political slant of the article. This suggests that the engaging nature of low-quality news is not solely driven by partisanship. The authors propose that factors such as novelty, negative emotional valence, and sensationalism likely contribute to this phenomenon.

The study also clarified the relationship between the volume of content and engagement rates. In terms of absolute numbers, users shared links to high-quality news sources much more frequently than they shared low-quality sources. High-quality news dominates the ecosystem in terms of prevalence. However, the engagement data indicates a discrepancy. While reputable news is shared more often, it generates less excitement or interaction per post compared to low-quality alternatives.

The inclusion of Mastodon in the dataset provided a significant control condition for the study. Because Mastodon does not use an engagement-based ranking algorithm to sort user feeds, the results from that platform suggest that algorithms are not the sole driver of the misinformation advantage. The fact that low-quality news still outperformed high-quality news on a chronological feed points to human psychology as a primary factor. Users appear to naturally prefer or react more strongly to the type of content found in lower-quality outlets.

But as with all research, there are some caveats. The data collection was restricted to a single month, which may not capture seasonal variations or behavior during major political events. The researchers were also unable to include data from Meta platforms like Facebook and Instagram, or video platforms like TikTok, due to data access restrictions. This means the findings apply primarily to text-heavy, link-sharing platforms and may not perfectly translate to video-centric environments.

Additionally, the study is observational, meaning it identifies associations but cannot definitively prove causation beyond the controls applied in the statistical models.

Future research directions could involve expanding the scope of platforms analyzed as data becomes available. Investigating the specific psychological triggers that make low-quality news more engaging remains a priority. The researchers also suggest that further work is needed to understand how the migration of users between platforms affects the spread of information. As the social media landscape continues to fracture, understanding these cross-platform dynamics will become increasingly important.

The study, “Divergent patterns of engagement with partisan and low-quality news across seven social media platforms,” was authored by Mohsen Mosleh, Jennifer Allen, and David G. Rand.

RELATED

Childhood ADHD traits linked to midlife distress, with societal exclusion playing a major role
Mental Health

Women who self-harm show altered brain responses to negative social media comments

May 25, 2026
Brain development patterns predict if childhood ADHD symptoms will fade or persist
Social Media

What happens when people get downvoted on Reddit? Scientists uncovered a surprising answer

May 23, 2026
TikTok tics study sheds light on recovery trends and ongoing mental health challenges
Political Psychology

TikTok disproportionately served anti-Democratic videos during the 2024 election, study finds

May 22, 2026
Liberals hesitate to share progressive causes framed with conservative moral language
Political Psychology

Liberals hesitate to share progressive causes framed with conservative moral language

May 18, 2026
Cognitive issues in ADHD and learning difficulties appear to have different roots
Sleep

Poor sleep and endless video scrolling form a predictable behavioral loop

May 17, 2026
Online trolls enjoy trolling, but not being trolled
Social Media

Americans systematically overestimate how many social media users contribute to harmful online behavior

May 14, 2026
Cognitive issues in ADHD and learning difficulties appear to have different roots
Mental Health

Taking a break from social media does not improve mental health, mass data review finds

May 6, 2026
Fascinating new neuroscience study shows the brain emits light through the skull
Social Media

Problematic social media use is linked to how feelings of freedom relate to mental health

May 5, 2026

Follow PsyPost

The latest research, however you prefer to read it.

Daily newsletter

One email a day. The newest research, nothing else.

Google News

Get PsyPost stories in your Google News feed.

Add PsyPost to Google News
RSS feed

Use your favorite reader. We also syndicate to Apple News.

Copy RSS URL
Social media
Support independent science journalism

Ad-free reading, full archives, and weekly deep dives for members.

Become a member

Trending

  • Being asked to help dampens the joy of doing good, according to children in multiple countries
  • Brain development patterns predict if childhood ADHD symptoms will fade or persist
  • TikTok disproportionately served anti-Democratic videos during the 2024 election, study finds
  • Neuroscientists discover the brain’s memory center starts “full” and prunes itself down to optimize learning
  • New study links manipulative personality traits to lower relationship intimacy expectations

Science of Money

  • What makes a public service job attractive? A new study sorts out which perks matter most
  • What a CEO’s tweets reveal about their paycheck
  • When optimism mutes the message: How investor mood shapes crypto’s response to economic news
  • Why nominal interest rates bite harder than textbooks suggest
  • California’s $20 fast food wage pushed restaurant prices up 3.4% across the state, new analysis finds

PsyPost is a psychology and neuroscience news website dedicated to reporting the latest research on human behavior, cognition, and society. (READ MORE...)

  • Mental Health
  • Neuroimaging
  • Personality Psychology
  • Social Psychology
  • Artificial Intelligence
  • Cognitive Science
  • Psychopharmacology
  • Contact us
  • Disclaimer
  • Privacy policy
  • Terms and conditions
  • Do not sell my personal information

(c) PsyPost Media Inc

Welcome Back!

Login to your account below

Forgotten Password?

Retrieve your password

Please enter your username or email address to reset your password.

Log In

Add New Playlist

Subscribe
  • My Account
  • Cognitive Science Research
  • Mental Health Research
  • Social Psychology Research
  • Drug Research
  • Relationship Research
  • About PsyPost
  • Contact
  • Privacy Policy

(c) PsyPost Media Inc