The next speaker at the 5th Data Donation Symposium at the Weizenbaum-Institut in Berlin is Simon Kruschinski, whose focus is also on the comparison of different data donation approaches. There are many different data donation approaches, and these need to be compared more directly to identify the least burdensome and most user-friendly frameworks.
This study seeks to compare download-upload and automated data donation approaches, to examine patterns in who donates data, and to determine how participants experience trust and privacy in data donation processes. It draws on an online panel of some 7,000 participants in Germany which is linked with digital behavioural data gathered from these participants.
Of these, the study focusses on some 2,000 Facebook or Instagram users in this panel, and asks them for their willingness to participate in data donations, with information on various personal and societal benefits and an offer of small monetary incentives; it also provides privacy and consent information, describes the kind of data to be donated, and outlines their data selection and deselection opportunities.
Once stepped through these stages, participants are then assigned to download-upload and automated data donation groups, where the data download requests are made; for the automated group this runs in a secure virtual environment where participants request their data from the platform, upload the data to a server, review their usage patterns, redact their data, and finally donate their data; for the download-upload group this uses the PORT data donation platform and involves several email remainders asking users to complete the donation process.
Finally, there will also be a post-donation survey exploring user trust and usability experiences with the process, which is also served to those who did not complete then donation process. All of this seeks to embed FAIR principles for the data donations.
Several practical and conceptual issues still need to be resolved here, especially also with the complicated nature of these processes, user concerns about data control, and cross-device usability. Feedback from users will be highly valuable here.
