The digital soup: the interplay between journalistic and algorithmic curation on news framing effects
OA Version
Citation
Abstract
The digital news ecosystem has evolved rapidly over the last several years. Driven by advances in artificial intelligence-mediated communication (AIMC) and algorithmic curation, these changes have transformed how audiences consume, perceive, and interpret news content on social media platforms and search engines. This dissertation introduces the concept of "digital soup" to describe these automated aggregation processes and explore the consequences of what happens when a diverse mix of news content types, sources and topics are sorted into a single information stream. Building on traditional framing theory, the study extends the concept of framing and its effects beyond individual news content to include the surrounding media blended into the digital soup. Specifically, “internal” framing devices represent classical framing techniques applied to news stories, while “external” framing devices denote peripheral content or other media that fill the broader context of news encountered in an algorithmic newsfeed. A unique contribution of this study lies in its focus on external algorithmic framing, operationalized through the surrounding context of posts such as memes and activism-based content. A 2x3x2 factorial online experiment is conducted to examine how these framing devices measured in the form of content types (humor, call-to-action) and different framing actors (journalistic, algorithmic, combined journalistic-algorithmic) influence users’ opinions of the news, their behavioral intentions, and their news credibility perceptions within a simulated social media newsfeed environment featuring two socially significant news topics: artificial intelligence (AI) and health. Findings revealed that both internal and external framing devices that emphasized urgency (call-to-action posts) and collective engagement (activism posts) produced stronger effects than humor and meme-based frames, particularly in health-related news. CTA framing increased policy support, while activism posts enhanced credibility perceptions of AI news. Significant covariate patterns highlighted the role of individual differences, reinforcing that framing outcomes, even in complex algorithmically-mediated environments, depend on how message tone and context align with audience dispositions.
This research presents a refined approach to understanding how algorithmic curation reshapes the traditional news landscape by influencing framing effects in heterogeneous digital information environments. By introducing and testing the concept of external framing devices, such as memes and activism posts that appear alongside news content, this study expands how framing is measured and theorized in relation to the algorithmically-curated newsfeed. Results showed that adjacent posts and the broader arrangement of content within a newsfeed environment play a critical role in shaping audience perceptions. This has direct implications for the design and governance of algorithmic platforms that distribute news. To support the democratic function of journalism in a rapidly evolving landscape shaped by AI and automated curation, platforms and policymakers should dig deeper into the contextual, surrounding elements in the design of algorithmic newsfeeds that affect how news is received, evaluated, and acted upon in the digital public sphere.
Description
2025