In a groundbreaking study conducted by researchers at Cornell University, an extensive analysis of millions of social media posts has revealed compelling insights into the dissemination of news content across various digital platforms. This comprehensive investigation sheds light on the persistent spread of lower-quality news, highlighting how partisan leanings and platform-specific user behavior interplay to influence engagement patterns. The study challenges preexisting notions about the dominance of any single ideological camp in driving the viral spread of questionable news online, instead unveiling a more nuanced, multifaceted landscape.
The research team set out to address a fundamental question: how does the quality of news content shared on social media correlate with user engagement across politically diverse platforms? Traditionally, much of the focus in this domain has been on Twitter (now known as X), primarily due to its open API and accessibility for academic scrutiny. Previous studies posited that right-leaning news outlets and users were disproportionately successful in garnering attention, especially for lower-quality or misleading information. However, Cornell’s latest research ventures beyond the confines of a single platform to examine seven diverse social media sites simultaneously, bringing a fresh and expansive perspective to the conversation.
The platforms selected represent a broad spectrum of political ideologies and community norms: BlueSky, Mastodon, LinkedIn, and Twitter/X were categorized as more liberal or politically neutral, while TruthSocial, Gab, and GETTR were identified with conservative leanings. By capturing every post containing links to news domains during January 2024—a total nearing 11 million posts—the researchers amassed an unprecedented dataset ripe for robust computational analysis. Their methodological rigor included isolating the effects attributable to content quality itself, independent of poster characteristics such as follower count or platform influence.
Central to the study is the operationalization of news quality. Instead of content-specific accuracy assessments, the team employed an established reliability rating system developed in 2023, which synthesizes numerous expert evaluations to assign credibility scores to over 11,000 news domains. This method offers a scalable proxy to estimate the veracity of linked news sources across millions of posts. Additionally, political orientation was ascertained and validated, allowing the researchers to correlate engagement trends with ideological alignment intrinsic to both the platform and the source material.
One of the most provocative findings is the consistent pattern whereby posts linked to lower-quality news receive higher engagement—measured through likes, shares, and comments—across every platform studied, irrespective of political stance. This phenomenon persists not only within algorithm-driven environments like Twitter/X and TruthSocial but also on platforms like Mastodon, which eschews ranking algorithms altogether. This strongly suggests that user preference, rather than automated amplification mechanisms alone, plays a decisive role in favoring sensational or less reliable news content.
The study nuanced earlier claims about ideological asymmetry by demonstrating that the amplification of low-quality news is not exclusive to the right-wing or conservative platforms. Although platforms with conservative user bases tend to share, on average, somewhat lower-quality news, the relative engagement advantage of such news over higher-quality content manifests equally on left-leaning sites. This “echo platform” effect, where conformity to the platform’s dominant political tone heightens engagement, underscores the role of identity affirmation and social reinforcement in shaping online information ecosystems.
Furthermore, the researchers controlled for numerous confounding variables to firmly attribute engagement differences to the content’s characteristics alone. This disentanglement from poster influence is critical, as it mitigates biases arising from highly followed users or influencers whose reach could otherwise skew results. By logarithmically adjusting for follower count and other user attributes, the study robustly demonstrates that quality deficits in news content, not the stature of the individual disseminator, predominantly account for increased virality on social media.
Beyond its implications for understanding digital information dissemination, the study holds significant import for public policy and platform design. The data signal a need to reconsider how social media ecosystems might inadvertently encourage the spread of unreliable information through inherent user preferences and social dynamics. Interventions that solely focus on algorithmic content moderation may be insufficient if user engagement patterns inherently favor sensational or partisan news regardless of accuracy.
David Rand, the lead investigator and professor whose expertise spans information science, marketing, management communication, and psychology, articulates the complexity of these dynamics. He compares social media interactions to “echo platforms,” where users affiliate with ideologically congruent content, and their engagement is heightened when such content aligns with the platform norm. This insight reconciles the paradox where misinformation appears virally potent across political divides, challenging the popular narrative that one ideological faction disproportionately drives misinformation spread.
The study also critically revisits the conclusions of the 2018 Science article that predominantly analyzed Twitter data and suggested particular political groups excelled at stoking false news engagement. By broadening the scope across multiple sites and incorporating political diversity among platforms, the current research refines and expands this earlier understanding, emphasizing the heterogeneity of social media landscapes and the limitations of generalizations drawn from single-platform studies.
Funding for this extensive research was provided by the Open Society Foundation, emphasizing the importance placed on unearthing empirical evidence to guide effective strategies in combating misinformation. The team’s work, published in the prestigious Proceedings of the National Academy of Sciences, exemplifies rigorous multidisciplinary scholarship. By merging computational social science with psychology and media studies, the research sets a new standard for future inquiries into the digital information age.
Cornell’s analysis offers urgent, data-driven insights to social media companies, policymakers, and users alike. Understanding that the problem of low-quality news engagement is endemic and ideologically widespread demands multifaceted responses that transcend partisan politics. Efforts to promote media literacy, incentivize high-quality journalism, and design platforms with nuanced user behavior in mind are crucial next steps supported by this landmark study.
This large-scale, methodologically robust analysis highlights the complex realities of modern social media engagement with partisan and low-quality news. It underscores the imperative to address the underlying human and systemic factors perpetuating misinformation in the digital sphere.
Subject of Research: Patterns of engagement with partisan and low-quality news across diverse social media platforms
Article Title: Divergent Patterns of Engagement With Partisan and Low-quality News Across Seven Social Media Platforms
News Publication Date: 30-Oct-2025
Web References: http://dx.doi.org/10.1073/pnas.2425739122
References: David Rand et al., Proceedings of the National Academy of Sciences, 2025
Keywords: social media, mass media, communications, social sciences
