Institute For The Future

Tell Me What I Can Do

 

Jay Parkinson has a typically smart post at the Future Well asking if "viewing personalized life data changes everyday behavior? For example, "he asks, "does Mint [the personal finance site] change an individual’s spending behavior for the better?" After noting that, at least at the moment, there's no definitive research that shows if looking at one's personal data can promote behavior change, particularly in health contexts, he offers up a good, though not exhaustive list of how data can be designed and visualized persuasively:

 

 

 

* Passive (the user has to do nothing to acquire the data)

* Non-invasive

* Real-time

* Focused (like a dashboard that measures only one thing such as fuel-efficiency, metabolic rate, or glucose)

* Linked in real-time to the desired effect

* Simple to gain insight and understand

* Linked to private, personal benefits (ex. decreased weight or increased mood)

* Linked to public benefits (ex. how decreasing caloric intake connects to global carbon footprint)

* Quirky, positive feedback

* Non-threatening negative feedback that doesn’t make you feel badly about your less than desirable behavior

* Socially connected to take advantage of human competitive nature?

 

Any more that need to be included?

 

 

 

 

I have a couple thoughts and additions. Data, without some simple steps to act on said data, is rarely helpful.

 

Take his example, Mint, which I've been using now for a couple years. I have made one significant change because of Mint--I transferred some money out of a zero-interest checking account and opened a high-interest savings account--something I had been meaning to do for several years but, as with many things in life, the short term hassle of going through the process kept me from doing so. Mint showed me the differences in effect over time--I'd have a specific dollar amount more in a year by transferring money to a savings account--and then it offered me a link to a bank where I could open up a savings account at that very moment. In other words, what made Mint persuasive wasn't simply a visualization of the current state of my finances, but:

1) Alternate visualizations/explanations of what my future personal finances could look like, based on my actions

2) A simple way I could take action immediately to put myself on the path to my preferred future.

 

This analogy to personal health only really works up to a point. I only had to act once to change my finances, but choosing the salad for dinner one time won't make me healthy. I need to choose the salad lots and lots of times, and manage to turn down the fries at least most of the time. That's a much trickier challenge.

 

Which gets to a broader question about what we mean by "data." I think he means the typical health stuff--weight, blood pressure, cholesterol, and so on. For those sorts of clinical markers, having specific, actionable recommendations about how to improve said markers is going to be hugely important.

 

Noting that there isn't any hard data on the effectiveness of Mint just yet, Parkinson asks:

 

 

 

But if Mint doesn’t significantly change people’s behavior, what will? And if Mint can’t do it with our money (everyone loves more money!), who will be able to do it with changing healthy behaviors (does everyone love more “health?”)

 

 

 

This last question--"does everyone love more 'health'"--is another way of getting at my question about the nature of data.

 

A couple months ago, I highlighted a New York Times article about a treadmill that, as one exercises, flashes images of foods corresponding to the number of calories burned--at 75 calories, you see a piece of tuna sushi, for example. At the time, I wrote "Tempting people with these indulgences in real-time, but only if they've earned it by burning off calories, offers a creative way for people to visualize pleasure and manage their health--and their indulgences--in effective ways."

 

Now, this is obviously not a foolproof concept. If, every time you go to the gym, you have a post-workout banana split, you're canceling out a lot of hard work. And more to the point, this only works if most of your other choices are healthy ones--if you already eat fast food three times a day, then go to the gym and see a tempting image on the treadmill screen, going for that post-workout banana split is an even worse plan.

 

But I do think there's something to thinking more creatively about what kinds of health data visualizations might promote behavior change. For most people, seeing their blood pressure drop a couple points isn't particularly persuasive--unless they already have hypertension. But everybody loves ice cream (or something like it.) And in that sense, I think the counterintuitive idea--that moderately unhealthy visualizations, such as a bag of chips or a dessert, that offer some clear guidance about when it would be okay to indulge in them, based on one's personalized data--might offer some possibilities to actually promote healthier behaviors.