Dougal Sutherland: Can mobile phone activity predict moods?

We’re constantly looking for ways to spot early signs of problems such as depression or bipolar disorder, based on the idea that the earlier we can spot signs the earlier we can intervene.

Traditionally in psychology this has relied on asking people to monitor themselves and report back. Lots of problems inherent in this e.g., lack of motivation, lack of awareness, not knowing what to look for, self-fulfilling prophecies, etc. Psychology research is often hampered by lack of more objective ways of collecting information that don’t rely so much on self-report of individuals.

New methods are emerging which look at how someone’s pattern of typing on a smartphone keyboard might predict early onset of depression.

Firstly – rest assured this isn’t some new and frightening emergence of big-brother watching your keystrokes! For this research people had to download a specific keyboard to use with their phone that feeds back results to researchers about their typing patterns, it didn’t capture the content of what they were writing. This is referred to as passive-sensing as doesn’t require any extra effort from people, just for them to keep using their phone as usual.

A number of interesting findings about depression:

• When depressed, people have lower accuracy in spelling and more variability in typing speed – possibly reflects slow-down in their thinking.
• They also tend to have more times using their phone than when not depressed – might indicate loneliness and social withdrawal.
• When people are depressed, they have much higher rates of using the backspace key than when not depressed – suggests they’re making more errors which could be due to slowing down of thinking or lots of ruminating about life and going round and round in their heads.

For people with bipolar disorder, they’re more likely to have high rates of backspace use when also having increased levels of activity and decreased levels of sleep – both of which are warning signs of a possible manic episode.

Other interesting research using technology:

Smartphone apps monitoring daily patterns of behaviour reduced reoccurrence of periods of depression. This was done by tracking when people slowed down in their levels of daily activity and had less periods of being outside in daylight. The app would alert them to this change, and suggesting that if the pattern continued, they could get depressed again – this led people to get more outside activity.

Some interesting possible implications:

• Could be really helpful in noticing early warning signs of things like depression in a really unobtrusive way.
• Imagine linking this data in with AI who could send you info about early warning signs and what to do about them, plus link you in with a psychologist if needed.
• Potentially gives people the ability to monitor themselves and make small changes in their behaviour to prevent getting depressed and without having to see a psychologist.
• Nice to have some good news about potential benefits of technology and given we usually focus on how these things can suck us down into the dark underbelly of the internet!

LISTEN ABOVE

See omnystudio.com/listener for privacy information.