My mobile, my doctor

Back in 2016-2017 I had a brush with something called Descriptive Experience Sampling, a technique for gathering information about what I call frames of consciousness, developed by one Russell Hurlburt. See, for example, reference 1. I have since wondered, on and off, how one might do better.

Then yesterday, in an email from the Kurzweil organisation, I was pointed at the 2015 study which can be found at references 2 and 3.

Leaving aside learned comments about statistics, surveys and methodology, it seems to me entirely likely that you could train a mobile phone to detect depression in the owner. In this case, it seems that the authors had a pretty good stab at it using no more than information about where the mobile phone was and about when it was being used, information which could be uploaded to the central database by app. In which connection, mobile phones have the great advantages of being connected, sophisticated devices which most people want to carry around with them, more or less wherever they are and whatever they are doing.

It occurs to me that it would not take much to add lots more to this basic information, with the consent of the owners, while avoiding intruding on the privacy of those with whom they are in contact, if not communication.

One could, without much difficulty, add information about stuff like air temperature, body temperature, blood pressure and heart beat rate. Indeed, I believe that there are plenty of joggers and worse out there who already do something of the sort.

One could analyse the speaker’s voice for signs of his or her emotional state. One could analyse word frequencies to the same end – which would not require the central computer to know anything about what was actually being said – which might worry the subject, the speaker.

One could do something with knowledge of the websites to which the mobile was connected. And for how long. Or perhaps the search terms which people feed into those websites.

Perhaps one could make use of knowledge of the micro-movement of the mobile, derived from the on board gyroscope, as apposed to the macro-movement, the location data.

One could certainly add something like Hurlburt’s random bleeper, to sample the subject’s state of mind. This would be something of a departure from the model so far, in that it would require the subject to actually do something, rather than just quietly collecting information in the background. But I am sure that there are plenty of people out there who would be happy to participate.

All in all, one could collect huge amounts of information about people, their state of being and their state of mind. And I imagine that there are plenty more people who would not mind this being done to them. Who might well regard it as contributing to science, as indeed it would be. Or at least, should be.

And if they checked the box for social worker, they would not be too surprised if a social worker turned up on their doorstep, worried about their state of mental or physical health.

Should we be worried?

PS 1: one comment about methodology might be the use of the one side of A4, PHQ-9 questionnaire at reference 4 to self-screen for depression. Does one really get useful results from such a tool?

PS 2: in the margins, I have learned all something called Craig's Lists, important as a listings service in the US, notionally not-for-profit. Something which, it seems, Judge Judy has known about for a long time.

References

Reference 1: https://psmv3.blogspot.com/2017/01/progress-report-on-descriptive.html.

Reference 2: Mobile Phone Sensor Correlates of Depressive Symptom Severity in Daily-Life Behavior: An Exploratory Study - Sohrab Saeb; Mi Zhang; Christopher J Karr; Stephen M Schueller; Marya E Corden; Konrad P Kording; David C Mohr – 2015.

Reference 3: https://www.egr.msu.edu/~mizhang/papers/2015_JMIR_MobileDepression.pdf. Open access source for reference 2.

Reference 4: http://www.cqaimh.org/pdf/tool_phq9.pdf. From Pfizer Inc. - who, I dare say, sell important anti-depressant potions.