A new study from Rutgers University researchers suggests that TikTok‘s content distribution patterns differ significantly from other social media platforms when serving content related to China, particularly regarding sensitive political topics.

The research, published in Frontiers in Social Psychology, combines three interconnected studies examining content distribution across social media platforms and user perceptions. The study analyzes how TikTok, Instagram, and YouTube handle content related to topics such as Tibet, Tiananmen Square, Xinjiang, and Uyghur rights.

First Study: Searches

In the first study, researchers created new user accounts across all three platforms to examine content served to users searching for these topics. The results show that TikTok consistently produces less content critical of the Chinese Communist Party (CCP) compared to Instagram and YouTube. For instance, when searching for content about Uyghurs, TikTok returned 10.7% anti-CCP content, compared to 84% on Instagram and 19% on YouTube.

The study also reveals that TikTok serves a higher percentage of irrelevant content for these searches. When users search for “Xinjiang,” 69.3% of TikTok results are irrelevant to the topic, compared to 6.7% on Instagram and 2% on YouTube.

Second Study: User Engagement

The second study analyzed user engagement metrics across platforms. Despite TikTok showing fewer anti-CCP posts, these posts receive significantly higher engagement when they do appear. On TikTok, anti-CCP content averages 113,767 likes per post compared to 28,151 likes for pro-CCP content. This pattern of higher engagement with critical content exists despite such content appearing less frequently in search results.

Third Study: Social Media Usage & Views on China

The third study surveyed 1,214 American adults about their social media usage and views on China. The research finds that increased time spent on TikTok correlates with more positive views of China’s human rights record and higher likelihood of viewing China as a desirable travel destination. This relationship remains statistically significant even when controlling for demographics and other social media platforms.

“Time spent on TikTok predicts beliefs about China’s human rights record over and above time spent on the other platforms and independent of user demographics,” the researchers write in their findings.

The study’s methodology included creating 24 new accounts on each platform, using U.S. IP addresses, and simulating teenage users. Researchers analyzed the first 300 videos returned for each search term, coding content as pro-CCP, anti-CCP, neutral, or irrelevant.

The research team acknowledges several limitations to their study. The simulated user accounts may not fully reflect the experience of long-term users who receive personalized content. Additionally, the sample for the survey portion, while large and demographically stratified, was not fully representative of the U.S. population.

Other Takeaways & Methodology

Key findings from the research include:

• TikTok produces nearly three times as much pro-CCP content as anti-CCP content, despite anti-CCP content receiving approximately four times more user engagement
• Users who spend more time on TikTok show a stronger correlation with favorable views of China (r=0.33) compared to users of other platforms
• The relationship between TikTok usage and favorable views of China remains significant even when controlling for age, political affiliation, and other demographic factors

The researchers note that while their findings suggest patterns in content distribution, they cannot definitively determine whether these patterns result from intentional algorithmic manipulation or other factors.

The study comes as TikTok faces increased scrutiny over its content moderation practices and potential influence on user perceptions. The platform, owned by Chinese company ByteDance, maintains that it operates independently of the Chinese government.

This research was conducted by Rutgers University’s Department of Psychology teams and the Network Contagion Research Institute. The complete study, including detailed methodology and data analysis, is available through the Open Science Framework.