harm
Abstract
Despite multiple reviews quantifying small associations between the time a person spends on social media and well-being, public and policy concern remains high, prompting calls for more nuanced approaches that incorporate both digital footprint data and content types. This study leverages second-by-second TikTok interactions to examine how specific social media behaviours and content consumption are associated with well-being. Analysing data donated by 161 TikTok users, we identify two distinct behavioural clusters: content creation (posting and commenting) and content consumption (watching, searching and liking). These findings challenge the notion that content consumption is a passive activity, revealing it as an engaged process influenced by user preferences. Notably, our results demonstrate that content type, rather than general usage metrics, shows stronger associations with well-being outcomes. Specifically, users with higher depression levels consumed more animals and wildlife videos, while those with higher anxiety levels consumed fewer travel and automotive videos (although these associations were very small). This cross-sectional study illustrates how granular, platform-specific behavioural data can refine the existing categorisation of social media use and suggests that content type warrants closer empirical attention than general usage metrics when examining associations with psychological well-being.
