Data Donation From YouTube and TikTok Users: How We Implemented It

Abstract

The rapid evolution of the Internet continues to reshape the methodological landscape of social science research. Digital methods, although relatively new, have already undergone several key transformations. From the initial ”follow the medium” paradigm, enabled by early 2000s website architectures and open APIs, to the current post-API context, marked by platform-driven restrictions on data access, researchers now face increasing constraints in obtaining platform-generated data. In response, a growing body of scholarship advocates for a shift toward user-centric approaches, with data donation emerging as a particularly promising methodological alternative. Data donation involves participants voluntarily sharing their digital trace data, offering rich, contextualized insights into online behavior while upholding ethical standards. Despite its potential, practical knowledge on how to implement data donation in empirical research remains limited.

This methodological article contributes to filling that gap by presenting the practical experience developed in the AlgoFeed project, in which 240 participants were involved in a study that included the donation of their YouTube and TikTok data to investigate feedback loops between algorithms and user behavior. We detail the project’s sampling strategy, data handling procedures, enrichment techniques, and legal considerations under the European regulatory framework.

This practical implementation illustrates how data donation can help overcome platform-imposed barriers while enabling more participatory and ethically sound digital research.