Research: TikTok Algorithm Shapes Teen Beliefs
New research explores how TikTok's algorithm doesn't just respond to teen preferences, but can actively shape them. The study suggests the platform's powerful recommendation engine can create echo chambers or fast-track polarizing content, influencing users' worldviews even through passive consumption.
The study is led by Munmun De Choudhury, a professor at Georgia Tech's School of Interactive Computing, with a multi-institutional team backed by a $1.7 million grant from the Huo Family Foundation. It uniquely analyzes passive consumption, using data archives shared by over 10,000 UK teenagers under GDPR regulations to see what the "For You" page shows them without active engagement. TikTok's recommendation engine uses machine learning to predict which videos a user will watch next, creating a personalized "For You" page for each individual. The algorithm analyzes signals like likes, shares, comments, and accounts followed, but a Wall Street Journal investigation found that the amount of time a user lingers on a single video is a particularly powerful factor in what content gets recommended next. Concerns about the algorithm's power are not new. A 2022 report from the Center for Countering Digital Hate found that new accounts registered as 13-year-olds were shown content related to self-harm and eating disorders within minutes of joining the platform. Research indicates it can take as little as 35 minutes of viewing for a user to form a habit on the app. Studies have shown the algorithm can rapidly lead users down rabbit holes of extreme content. One auto-ethnographic study created a moderately conservative profile and found that after watching about 1,000 videos, the recommendations became increasingly radical, leading to conspiratorial echo chambers. Another report noted a fourfold increase in misogynistic content recommendations after just five days of light engagement with such material. The speed at which the algorithm personalizes content has been described as creating a "closed-loop digital ecosystem." This system can inadvertently reinforce certain beliefs or expose vulnerable teens to harmful ideologies related to everything from body image to political extremism. The process has been termed the "TikTok-ification" of extremist material, making it more appealing through short, aesthetic videos.
