Collaborative Project SCREEN: DFG Funds Data Privacy Compliant Research on Smartphone Usage

How do people actually use their smartphones, and what kinds of content do they encounter in apps, social networks, or messaging apps? Above all, however: How can this be studied without compromising data privacy? These questions are the focus of the collaborative project “SCREEN – Smartphone Content Research EcosystEm for TrackiNg,” which will be funded by the German Research Foundation (DFG) starting January 1, 2027.

Participants include ZeMKI member Prof. Dr. Cornelius Puschmann from the University of Bremen, Prof. Dr. Philipp Krieter from Düsseldorf University of Applied Sciences, as well as Dr. Lukas Rahnke-Otto and Dr. Mareike Wieland from GESIS. The research project is scheduled to run for 36 months.

A New Approach

Previous research on smartphone use has relied primarily on surveys, but the self-reported data from these surveys is only accurate to a limited extent. While tracking methods provide more detailed data, they are often limited to individual applications or collect large amounts of sensitive information.

SCREEN therefore takes a new approach: research-relevant data is processed directly on the smartphone. Private screen recordings do not leave the device and are deleted after processing. For scientific analysis, only anonymized or pseudonymized information required for the respective research question is made available.

Technical Implementation

The technical foundation is an approach already developed by Prof. Dr. Krieter and other researchers that automatically analyzes screen activities. Using machine learning and automated text recognition, a module is designed to identify specified terms or topics across various apps and media formats. At the same time, a second module tracks which apps are used and for how long. This means the focus is not only on screen time but also on specific content and the context of use.

Pilot Study

The pilot study examines news consumption: tracking data is combined with mobile surveys to compare actual usage with users’ self-perceptions. This allows researchers to investigate when and how news is consumed, how well-informed users feel, and how news consumption relates to well-being. At the same time, the pilot study serves as a field test of the software under real-world conditions.

Goals

• Integration of the SCREEN modules into the open-source tool GESIS AppKit for mobile surveys
• Tests with researchers and students to assess the software’s user-friendliness
• Evaluation of whether AI language and multimodal models optimized for smartphones can classify content directly on the device – so that sensitive data is processed on the end device

In the long term, SCREEN is intended not only to answer specific research questions but also to create an infrastructure for future studies in the social and behavioral sciences.

Further details on the research project “SCREEN – Smartphone Content Research EcosystEm for TrackiNg” can be found in the article by the Düsseldorf University of Applied Sciences.