I recently ran across two different, new apps in development for smartphones and iPhones, both of which purport to measure a person’s mental health, happiness and even depression completely passively. (“Apps” are tiny pieces of software that run most commonly on portable devices.)
This, of course, is a Big Deal, since one of the major stumbling blocks of the thousands upon thousands of health apps are their need for something or someone to input personal health data. Without personal health data, health and mental health apps are generally pretty useless.
The method to measure one’s psychological well-being (or, as we more commonly refer to it, one’s happiness) passively is to use whatever metrics are available through the phone. Since phones generally only have a limited amount of inputs — voice, video, geo-positioning (GPS), and an accelerometer — your choices as a researcher interested in personal health data are pretty limiting.
Using only these four physical measurements, is it really possible to accurately and reliably measure a person’s well-being? Let’s find out.
In the research I’ve read on this issue, the researchers focused on three components: social interaction, as measured solely by the amount of talking done (through the phone’s microphone); sleep/wake patterns (through the accelerometer); and physical activity (through the accelerometer and geo-positioning).
Let’s look at each one of these characteristics in turn.
1. Social Interaction as Measured by Talkativeness.
Slowed speech or a reduction in the amount of speech (from whatever the person’s original baseline was) can be one part of one symptom (of the 9 symptoms) of depression. It is, however, also a symptom of many, many other disorders. Slowed speech, or a reduction in the amount of speech is part of a larger symptom cluster in depression, called “psychomotor agitation or retardation.” The DSM-IV makes clear that this slowed speech or reduction in the amount of speech can’t just be a subjective feeling — it has to be severe enough to be observable by others.
It’s also important to note that since people vary widely in terms of their social outgoing-ness (extroversion) and talkativeness, anything that seeks to measure how much a person is talking throughout the day is going to have to understand that individual’s personal talkativeness baseline level.
For instance, if I usually say about 20 sentences a day, and then I go down to 10, that might be an important change. But the app would have to know my baseline first. If it just assumes that I’m like an average person who says 200 sentences a day (or whatever the real number is) and sees I’m not meeting that average, it’s going to be wildly inaccurate.
Last, the most obvious problem with trying to measure social interaction or isolation through sound alone is the reality of how we conduct ourselves through technology. Much social interaction done today is done silently, through our smartphones and keyboards. It also assumes that simply sitting in the same room quietly with another person is the same as sitting alone in your own room. Being together with others, but not necessarily talking, is the New Togetherness.
Researchers can also look at “stress levels” in one’s speech. I suppose that could indeed give you an immediate, real-time reaction to things happening in the world around you. But good mental health isn’t based upon simply your levels of stress — it’s based on how resilient you are and what you do with such stress later on. These are vital components a smartphone or iPhone simply can’t measure.
2. Sleep/wake patterns.
Problems with sleeping aren’t going to be detected by the app, since it can’t tell when you’re sleeping or not (unless you pick up your smartphone every time you wake up). What it can do right now within 1 and 1/2 hours of waking is to determine if you’re awake or not (because you start using your phone). An hour and a half is a huge degree of error, and can easily be the difference between you getting a normal night’s sleep (8 hours, say) and not (6 1/2 hours).
The sleep/wake cycle is also impacted by dozens of other variables that may have nothing to do with your overall well-being or happiness. These include things such as a change of season, change of working times, change in relationship status, change of child rearing duties, change in physical health, and about a half dozen other mental health concerns.
It could also include a new exercise routine, getting married, or moving into a new house — all things that most people would think of as positives and increase happiness. Yet the app would see them as negatives, since they all might impact your normal sleep schedule.
An impact in your sleep schedule is not really a sufficient indicator of much of anything — other than you have trouble sleeping. It could be caused by so many different things as tying it to just one thing is simply not very reliable — or scientific.
3. Physical activity.
Physical activity is correlated positively with increase happiness, as well as greater overall health. People who engage in regular physical activity indeed may feel better about themselves and have a better mood.
In fact, if there’s one thing you want to do today — right now! — to make yourself feel better, go take a walk.
But a lot of people’s physical activity is pre-determined by the type of work they do and the lifestyle they lead. So if you’re tied to a desk all day, chances are your physical activity measurements are always going to be worse than someone who works outdoors all day. Even if you work-out on a regular basis.
Physical activity alone is not really a good measurement of mood. And while someone who engages in more physical activity should be at lesser risk — population-wise — for depression or other mental health issues, it can’t speak at all to an individual’s risk. After all, a professional athlete who is engaged in physical activity almost every day can still become depressed.
Can an iPhone Measure Your Mood?
Which brings us back to the original question… In reviewing what we know about mental disorders, depression and happiness, is any smartphone or iPhone app really going to be an accurate measure of those things?
Probably not. While researchers may find some weak correlations with some of these things and mood, I have my doubts about whether such an app can be robust and personalized enough to actually give most of us meaningful information.
So do we need an app to tell us we’re depressed? Most of us are already quite well enough aware of when we’re feeling down, socially isolate, or don’t feel like talking to others.
And an app is especially ironic, given the active initiative you would have to take in order to use it. You would need to download and install the app first — suggesting you already have a certain amount of insight into your own mood or psychological needs.
For further reading…
- Get Some Therapy From An App That Reads Your Feelings Through Your Voice
- Web-based counseling — Telepsychiatry — is taking off