The ubiquity of smartphones and the data collected from them may be an important tool in understanding human behavior. Studying human psychology depends on constant observation, but it would be time-consuming, expensive, and intrusive for a researcher to “follow people around 24/7 with a notepad,” says Sandrine Müller, a computational social scientist and postdoctoral research fellow at the Data Science Institute at Columbia University.

“Smartphones are a psychologist’s dream come true as they allow us to collect fine-grained behavioral data in a non-intrusive, inexpensive, and continuous manner,” she says. “It allows us to use the big data produced by people’s smartphones to improve their well-being—a data-driven method that’s helping to transform the fields of behavioral science and human psychology.”

Research shows that depression can be marked by a change in behaviors, and the mobility data could allow the authors to detect indicative behavior changes in participants. People experiencing depressive episodes, for example, might start to leave home less often, socially isolate themselves, visit fewer places, and become less physically active.

In a recent study, Müller and collaborators from Columbia Business School collected and monitored the mobile phone GPS data of 2,319 students enrolled in a psychology course at the University of Texas at Austin. During two-week tracking intervals in the fall semester of 2016 and the spring semester of 2017, the students were sent sampling surveys four times a day between 12 p.m. and 9 p.m. about their moods and mental states. The team also tracked the students’ hourly movements and documented the places they visited, their routines, and the distances they traveled.

Combining the survey data with the mobility data, Müller, Heinrich Peters, a doctoral candidate in management at Columbia Business School (CBS), and Sandra Matz, an assistant professor of management at CBS, made connections between a student’s sense of well-being and her or his mobility. Their findings, which will be published by The European Journal of Personality in a special issue devoted to behavioral personality science in the age of big data, suggest that students who reported being more anxious or more stressed traveled shorter distances; that students who were more depressed or lonely showed more similar patterns across days; and that those who spent more time in social places felt less lonely.

“We were delighted with the results of our study and hope the techniques can be used, for instance, to help study the patterns of people who are depressed, which might look very different for different people,” says Müller, who is an alumna of Dresden University of Technology and the University of Cambridge. “Psychologists could then reach out to them with interventions that can help them feel better.”

This research also breaks new ground, Müller says, in that previous work on the connection between mobility and psychology have looked at very specific aspects of the two areas and have been inconclusive. The new study, conversely, documented the connection by examining two types of mobility behaviors—the students’ movement patterns and the places they visited—and linked that data to six indicators of well-being: depression, loneliness, anxiety, stress, affect, and energy.

— Robert Florida