Identifying Political Manipulation Behind TikTok's Algorithm

Last registered on April 23, 2026
View Trial History

Pre-Trial

Trial Information

General Information

Title
Identifying Political Manipulation Behind TikTok's Algorithm
RCT ID
AEARCTR-0018374
Initial registration date
April 15, 2026

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
April 23, 2026, 9:24 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Primary Investigator

Name
Kuan-Ming Chen
Affiliation
National Taiwan University
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
C. Y. Cyrus Chu
PI Affiliation
Institute of Economics, Academia Sinica
Contact Investigator
PI Name
Ming-Jen Lin
PI Affiliation
Department of Economics, National Taiwan University
PI Name
Ming-Chi Chen
PI Affiliation
Institute of Sociology, National Tsing Hua University
PI Name
Chien-Yu Lai
PI Affiliation
Department of Economics, National Taiwan University
PI Name
Yu-Chang Chen
PI Affiliation
Department of Economics, National Taiwan University
PI Name
Greg Chih-Hsin Sheen
PI Affiliation
Institute of Political Science, Academia Sinica
PI Name
Chan-Hsi Wang
PI Affiliation
Institute for National Defense and Security Research

Additional Trial Information

Status
In development
Start date
2026-05-01
End date
2026-12-31
Keywords
Behavior
Additional Keywords
Social media censorship, Algorithmic manipulation, TikTok, Content moderation, Field experiment, Information control, Political censorship
JEL code(s)
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
This study investigates whether TikTok's recommendation algorithm systematically suppresses content that is politically sensitive to the Chinese Communist Party (CCP). In the current trial, we use 200 mobile phones, each configured with a unique TikTok account, to upload 13 distinct text-based video stimuli—each uploaded 100 times across devices—for a total of 1,300 uploads. The stimuli are paired treatment/control texts (in Chinese) spanning five thematic domains: (1) "Wuhan Pneumonia" (武漢肺炎) vs. "COVID-19," (2) "Falun Gong" (法輪功) vs. "Tai Chi" (太極拳), (3) "Support Taiwan Independence" (支持台灣獨立) vs. "Taiwan Independence" (台灣獨立) vs. "Oppose Taiwan Independence" (反對台灣獨立), (4) "Support Palestine Independence" (支持巴勒斯坦獨立) vs. "Palestine Independence" (巴勒斯坦獨立) vs. "Oppose Palestine Independence" (反對巴勒斯坦獨立), and (5) "East Turkistan" vs. "West Turkistan" vs. "Xinjiang." Sensitive text is embedded in TikTok posts using a graphic overlay technique so that only TikTok's algorithm—not human viewers—can identify the text content. The primary outcome is the play count of each video measured at multiple intervals after upload (up to 12 hours).
External Link(s)

Registration Citation

Citation
Chen, Yu-Chang et al. 2026. "Identifying Political Manipulation Behind TikTok's Algorithm." AEA RCT Registry. April 23. https://doi.org/10.1257/rct.18374-1.0
Sponsors & Partners
Experimental Details

Interventions

Intervention(s)
Intervention Start Date
2026-05-01
Intervention End Date
2026-12-31

Primary Outcomes

Primary Outcomes (end points)
Play count (views) at 12 hours post-upload: The number of views each uploaded video receives within 12 hours of posting, collected via automated web scraping.
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
Play count at intermediate intervals: Views measured at 0.1, 0.5, 1, 3, and 6 hours after upload.
Like count at 12 hours post-upload: The number of likes each video receives within 12 hours.
Comment count at 12 hours post-upload: The number of comments each video receives within 12 hours.
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
The whole experimental design is described in the hidden part.
Experimental Design Details
Not available
Randomization Method
Randomization is done by a computer.
Randomization Unit
Each of the 13 video scripts is uploaded approximately 100 times, yielding 1,300 total uploads.
200 mobile phones are used; each phone uploads one video per half-day session.
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
N/A (not clustered)
Sample size: planned number of observations
1,300 (13 video scripts × 100 uploads each)
Sample size (or number of clusters) by treatment arms
Each of the 13 video scripts is uploaded approximately 100 times, yielding 1,300 total uploads. For a detailed breakdown of the unit, see the experimental design.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
We estimate that 100 uploads per arm provides sufficient statistical power (≥ 0.80 at α = 0.05) to detect suppression effects of the magnitude. We use 200 phones with tighter control over device-level heterogeneity.
Supporting Documents and Materials
IRB

Institutional Review Boards (IRBs)

IRB Name
National Taiwan University
IRB Approval Date
2026-04-08
IRB Approval Number
202603HS028
Analysis Plan

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information