Research Summary: Participant-Mediated Collection of Sensitive Digital Trace Data: The CANDOR Research Infrastructure

The research outlines a novel approach, 'participant-mediated data donation,' for collecting sensitive digital trace data. This method allows individuals to contribute selected portions of their digital histories, addressing limitations of traditional data collection, such as declining platform API access and reliance on publicly observable data. It enables the collection of longitudinal, non-public, multi-platform data, offering richer insights but also introducing methodological challenges related to participant control and data minimization.

Why it matters

This development is strategically important as it proposes a new paradigm for data collection, circumventing existing technological and access limitations. It unlocks opportunities for deeper insights into behavior by accessing previously unavailable or difficult-to-obtain data, which can inform product development, policy, and societal understanding.

Key insights

• Digital trace data offers detailed behavioral measures in everyday environments.
• Traditional data collection methods are constrained by reduced platform API access and a historical focus on public data.
• Participant-mediated data donation is an alternative where individuals volunteer specific parts of their digital history for research.
• This method facilitates the collection of longitudinal and non-public behavioral data across various platforms and modalities.
• It allows for linking donated data with independently collected study measures, enabling new research designs.
• The approach presents opportunities for richer research but also introduces methodological challenges, particularly concerning participant control and data minimization.

Source

arXiv — Computers and Society — https://arxiv.org/abs/2609.17722