Information Silos on Social Media

How to cite

Han, Tianyu; Wang, Wenbo; Shi, Zijun. 2025. "Information Silos on Social Media." AEA RCT Registry. AEARCTR-0017569. https://doi.org/10.1257/rct.17569-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Tianyu Han* , Wenbo Wang , Zijun Shi

* Primary investigator

Registry ID
AEARCTR-0017569
DOI
https://doi.org/10.1257/rct.17569-1.0
First registered
Last updated
Status
On going
Trial dates
2025-12-30 – 2026-01-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/17569

Abstract

In the digital age, social media platforms are central to the exchange of information. Despite the vast accessibility of content, users often find themselves in information silos where a small set of similar interests and ideas are reinforced, limiting exposure to diverse perspectives. This phenomenon raises concerns about its potential to exacerbate societal divisions and influence political, social, and economic dynamics. Our research aims to systematically measure the presence and intensity of these information silos on social media platforms. We will assess how consumers react to these silos, including their awareness and behavioral responses, and explore various interventions designed to mitigate the effects of these silos.

We introduce a novel Information Silo Index designed to evaluate the extent of content diversity or limitations encountered by users. This index is built on two key components: intrapersonal variety, which assesses the range of topics a user engages with, and interpersonal difference, which measures how a user's content consumption differs from the broader population.

Our study targets TikTok users and employs a combination of surveys and experimental methods to gather data. Participants will complete demographic surveys and record their TikTok browsing behavior. By analyzing video topics, we will compute the Information Silo Index and correlate it with user characteristics. This approach will help us identify patterns and trends in silo formation and assess consumer awareness and reactions to these silos. We will also implement information treatments to explore how users respond to messages about their browsing habits. By providing tailored feedback, we aim to disrupt existing silos and encourage broader content engagement.

Through understanding the mechanisms behind information silos on social media, we hope to provide actionable insights for both users and platform designers, ultimately contributing to a healthier public discourse and a more informed society.

Intervention

Different participants will see different treatment messages about the topics they viewed on TikTok

Experimental design

We recruit US participants on Prolific, ask them to submit TikTok video captures, and label the topics of those videos to measure information silos on social media. Then, we distribute information treatment and evaluate how consumers respond to these interventions.

Design details

Randomization method
Randomization by a computer
Randomization unit
Individual
Clusters
2000 individuals
Observations
2000 individuals from Prolific
Arms
400 per treatment arm
Primary outcomes
Measure of information silo indices

Cited by

No citation data available for this registration yet.

Keywords