Digital Phenotyping – Lessons Learned, Challenges, and Future Perspectives for Psychodiagnostics

Abstract: Digital phenotyping uses data collected from smart devices to track and predict biological, psychological, and social aspects of human behavior and experience. This approach shows great potential for psychodiagnostics. This article describes possible applications for psychodiagnostics. This study draws on empirical insights from the Corona Health App study, an initiative conducted during the COVID-19 pandemic in Germany, to examine the lessons learned. The Corona Health App study recruited a large number of participants ( N = 4,041) over a broad age range. However, the predictive utility of the sensing data was limited compared to self-report measures and findings from other relevant studies, potentially because of short observation windows, low sensor richness, and a focus on between-person variation. These limitations point to the need for richer, longitudinal data collection and more nuanced intraindividual analyses. The project also highlighted the importance of interdisciplinary collaboration and preregistration and reporting standards. Based on these findings, the paper offers practical recommendations for advancing digital phenotyping research.