Network Contagion Research Institute Intelligence Report 23 LIMITATIONS While our study provides significant insights into the manipulation of content on TikTok and other social media platforms, there are inherent limitations that should be acknowledged. First, our research primarily relies on the analysis of content served to newly created, organic accounts. While this methodology is designed to mimic the experience of typical users, it does not account for personalized content that may be delivered based on individual user histories and interactions over time. Consequently, the data may not fully capture the breadth of content suppression or amplification experienced by all users. Additionally, the coding and classification of content as pro-China, anti-China, or irrelevant involved subjective judgment, even with the use of blind coding and a third-party arbitrator for discrepancies. Although efforts were made to minimize bias, the potential for interpretative differences remains. Furthermore, our study did not explore the full range of user engagement metrics, such as comments and shares, which could provide deeper insights into user reactions and the spread of content across networks. Lastly, a clear limitation in this report, which compared video streaming platforms, was that it was not a fully comprehensive cross-platform analysis. Our findings highlight a novel and concerning shift from merely observing evidence of content influence on TikTok to detecting psychological changes among its users because TikTok users not only encounter biased content but as hypothesized, also exhibit altered perceptions and attitudes in favor of the CCP. More comprehensive efforts are needed to better understand the relationship between user psychology and the prevalence of manipulated content.