Sensor technologies, place, and pattern in mental health research

• In Event: Well Years, Good Years, Quality Years – Calibrations and Aggregations of Daily Living: Temporalities

Abstract

This paper considers a trend in mental health research to use passive sensor data to detect mental health symptoms. In this model of detection “algorithms will be used to estimate new measures of mental state and behaviour based on digital data” (Glenn & Monteith 2014). Sensor apps can be used to detect activity, location (and therefore routine), and social connectivity. These emerging technologies represent a potentially powerful approach to predict mental health risk.

Various terms have been used to describe the use of ubiquitous sensor data to estimate behaviours, such as reality mining (Eagle & Pentland 2006), personal informatics (Li et al. 2010), digital phenotyping (Jain et al. 2015, Torous et al. 2016), and personal sensing (Klasnja et al. 2009). Each of these different terms carries with it an imaginary of what data can do or of what can be done with data.

My discussion here is concerned with these data imaginaries, as well as the question of what it means to determine risk on the basis of pattern or routine. What does it mean to fit to a pattern of risk or against a pattern of risk? Is it your deviation from an expected path that may flag you as “at risk”? Or does the creation of the data about ones everyday life change the patterns? What kind of effects will these flags or markers have on our lives, if this kind of analysis becomes the norm?

Author

• Frances Shaw, Black Dog Institute