Extracting a digital sociability score from passive smartphone data

Activity: Talk and presentationAcademic presentationAcademic


Social interactions are a fundamental part of human healthy life. Social dysfunction and social withdrawal are associated with higher mortality and morbidity risk, including cardiovascular diseases, and are one of the first and most common signs of major neuropsychiatric disorders like Schizophrenia and dementia.
Social functioning is traditionally measured using questionnaires. Though, questionnaires require active participation of patient and interviewer, are difficult to scale to large number of patients, are limited to only few timepoints, and suffer from non-objective answers by the participants.
Smartphones are nowadays very widely used, and are a powerful tool to measure (social) behaviour of people. We aim at exploiting smartphone usage data (i.e. no content of messages and calls, but only phone usage and location data), that are passively collected by the smartphone application Behapp (behapp.com), to objectively measure social behaviour. This measurement technique has the advantage of not requiring any action by the patient after app installation, it can be easily scaled to large number of patients, and it objectively and continuously measures (social) behaviour.
The Behapp application has been already successfully used for assessing neuropsychiatric phenotypes.
It also successfully measured behavioural changes during the COVID-19 pandemic, where imposed quarantine restrictions resulted in more time spent using communication apps, more time spent at home, and less number of places visited.
The continuous measurement and tracking of (social) behaviour over time can be exploited to detect early warning signals (in our case social withdrawal) to successfully tackle the onset or the relapse of psychiatric diseases. Yet, a major challenge to detect early warning signals with smartphone data is to convert raw smartphone data into a measurable quantity describing social behaviour. Extracting such a digital sociability score, and validating it with questionnaire data as reference, is the aim of the present work.
Event titleEuropean Conference on Behavioural Biology 2022: All of life is social!
Event typeConference
LocationGroningen, NetherlandsShow on map
Degree of RecognitionInternational