User-centric approaches for collecting Facebook data in the ‘post-API age’: experiences from two studies and recommendations for future research

Although other social media platforms have seen a steeper increase in users recently, Facebook is still the social networking site with the largest number of users worldwide. A large number of studies from the social and behavioral sciences have investigated the antecedents, types, and consequences of its use. In addition or as an alternative to self-reports from users, many studies have used data from the platform itself, usually collected via its Application Programming Interfaces (APIs). However, with the drastic reduction of data access via the Facebook APIs following the Cambridge Analytica scandal, this data source has essentially become unavailable to academic researchers. Hence, there is a need for different modes of data access for what Freelon (2018) has called the ‘post-API age’. One promising approach is to directly collaborate with platform users to ask them to share (parts of) their personal Facebook data with researchers. This paper presents experiences from two studies employing such approaches. The first used a browser plugin to unobtrusively observe Facebook use while users are active. The second asked participants to export and share parts of their personal Facebook data archive. While both approaches yield promising insights suitable to extend or replace self-reports, both also entail specific limitations. We discuss and compare the unique advantages and limitations of both approaches and provide a list of recommendations for future research.