You may think you know a lot about the inner workings of your phone, but your phone may have an even better picture of your mental state.
A new study published yesterday in the Journal of Medical Internet Research by researchers from Northwestern University details how GPS and usage sensors can determine whether or not a user is exhibiting signs of depression. The report detailed that researchers were able to use the data from smartphones to determine a participant’s feelings of depression with 86.5 percent accuracy.
The study tracked 40 participants who carried phones with a sensor data acquisition app installed, called Purple Robot, that describes itself as a “sensing and scripting application that enables the creation of context-aware behavioral interventions and experiences.”
The report details that the researchers then gauged which participants in the group had feelings of depression through a self-reported questionnaire:
At the beginning of the study, participants were asked to complete an online assessment consisting of a demographics questionnaire and the Patient Health Questionnaire-9 (PHQ-9), a commonly used measure of self-reported depressive symptom severity , which produces scores ranging from 0-27. Scores of less than 5 indicate no depression, 5-9 mild depression, 10-14 moderate depression, 15-19 moderately severe, and over 20 severe depression [ ].
From here, researchers analyzed the data, tracking how often users were on their phones, how often they were leaving home, and how much they were generally deviating from their daily route. They found, unsurprisingly, that users “with depressive symptoms tend to move less through geographic space.”
In the report’s conclusions, the report’s author, David C. Mohr, a Professor in Preventive Medicine-Behavioral Medicine at Northwestern University, discussed the implications of the study for patient care in “unobtrusively” detecting behavioral factors that may be related to depression.
This study obviously brings up some pretty clear privacy concerns, especially when GPS coordinates and log-in data are classified as metadata, which is what the government has been ‘legally’ siphoning through its bulk data collection of Americans’ cell phone calls for years. Purple Robot is definitely way more sophisticated than a couple of recurring data points, but this research really does display what results the right analysis can reach when it knows where people are and how they are interacting with technology, things your phone has a pretty decent handle on.