Problematic smartphone usage affects many people. Even behavior that is not initially intentional, but which then leads to unwanted prolonged scrolling on the smartphone (known as “doom scrolling” or the “rabbit hole” trap), is considered problematic smartphone usage. According to estimates, 1-2% of the German population has been diagnosed with an internet use disorder, and approximately 5-7% are classified as problematic users. However, the number of unreported cases is likely to be much higher. 

Problematic smartphone use can lead to health problems such as nervousness, unproductivity, depressive moods, concentration problems, exhaustion, and fatigue. 

However, it is difficult for users to recognize when their own behavior becomes problematic. Existing screening methods such as self-monitoring apps or questionnaire screenings can provide guidance, but they usually only capture quantity or subjective assessments, not the objective quality of use. In the KiQuNu project, we therefore want to explore low-threshold screening using AI that utilizes real usage data, recognizes problematic patterns, reveals trends, and answers the question for users: “How problematic is my cell phone usage?”

The aim of this feasibility study is therefore to find out whether patterns can be identified in real smartphone usage data with the help of AI that can serve as a basis for an objective, reliable, and valid assessment of the quality of smartphone usage and as the basis for a self-monitoring app.

Together with our partner, the start-up Detoxi, we are first developing a voluntary data donation app. This is used to collect real context-specific data, such as time, screen time, etc. We use this to develop AI algorithms to discover recurring individual patterns and derive problematic smartphone usage behavior from them. This information is then reflected back to the user in an app as feedback, and potential risks are assessed.