Digital Phenotyping in Psychology: A Quantum Leap for Psychological Research?

Abstract

Digital phenotyping constitutes a new and powerful psychodiagnostic approach in many areas of psychology and medicine. The basic idea behind digital phenotyping is using digital traces gathered in everyday life to check their predictive potential. Digital traces may reveal predictive power capabilities for a wide range of applications. The prerequisites for a successful implementation are elaborated smart sensing approaches, big data-based extraction (data mining), and machine learning-based analysis methods. Initial empirical studies illustrate the high potential but also the research methodological as well as ethical and legal challenges to establishing reliable findings beyond random correlative results. In particular, legal and ethical guidelines need to set the siderails, ensuring that the findings are utilized in ways desirable to both individuals and society. To conclude, digital phenotyping offers a wide range of opportunities for psychology as a research and teaching domain, which, on the one hand, strives for lived collaboration between specialized fields, and, on the other hand, wants curricular extensions. The present narrative review provides a theoretical and nontechnical introduction to the research field of digital phenotyping, initial empirical findings, a discussion of its possibilities and limitations, and recommendations for action areas.