Self-reported and logged time spent on TikTok: The role of engagement and use fragmentation

Researchers leverage information on the amount of time people spend using digital media for various purposes, such as understanding the impacts of over-usage on mental health. Time spent on digital media is commonly measured by self-reports, in which participants estimate their own usage. However, prior work has identified systematic inaccuracies in self-reports of time spent on social media when compared to logged usage data. In this study, we build on prior work by examining TikTok, a social media platform centered around short-form video content, in the context of self-reported time estimation. We conduct platform-independent measurements of self-reported and server-logged TikTok usage time (n = 255) to examine how engagement behaviors shape users’ perceptions of time spent on the platform. Our results show that participants systematically overestimate their time spent on TikTok, and that higher levels of engagement, particularly liking more videos, are associated with greater estimation error. We identify engagement factors correlated with self-report inaccuracies and extend prior work on time estimation across digital media contexts.