Longitudinal Data Donation Behavior and Data Omission across Four Social Media Platforms

Abstract

This research note presents novel insights from a two-wave, longitudinal data donation study encompassing four major social media platforms: TikTok, YouTube, Instagram, and Facebook. We delve into a critical, yet underexplored, aspect of data donation research: the ethical provision for participants to delete any data traces they are uncomfortable sharing within standard Data Download Package (DDP) collection pipelines. Our analysis provides cross-platform insights into how user-side coverage error, stemming from such data trace omission, impacts the quality of DDPs and, consequently, the analytical power of the collected data. Crucially, our longitudinal design enables an examination of whether and how participants alter their data donation and deletion behaviors over time, shedding light on the stability of these choices in a panel setting. Our findings reveal an overall platform donation rates increase in the second wave. Omitting data traces is predominantly a characteristic of first-time donors, with very few participants omitting data in both waves, suggesting an increased trust among repeat donors.