Every day you interact with the web. You log on. You upload, you download. You tap and you click. You search, you “like”, you pin, and you retweeet. These actions make the web work for you, but they also make you work for the web. It should come as no surprise to even the casual technology observer that we are now living in the age of data. Some call it “big data”, but instead of thinking about it as a thing, we can also think of it as a an ecosystem that can be described by its fundamental structure – the database. Our lives and the actions we engage in on a daily basis are constantly being accessed and stored in a database. Our actions may be passively collected (think about how Google’s Adsense operates) or actively collected (checking in on Foursquare or updating Twitter). While it may seem as if we are living and engaging with a dystopian ecosystem, we believe that there are possibilities for engaging and enhancing our current health experiences by taking advantage of our personal and social databases.
We don’t need to rehash the idea that we are also in the midst of an explosion of tools and services that support the gathering of health-related data. If you’re reading this, you know that the Quantified Self movement is gaining traction and new devices and applications are being introduced at a rapid rate. Naturally, these tools are heard towards helping an individual lead a healthier life. This inherently creates a future-focus environment in which the user is presented data, analytics and recommendations for positive health behavior change in the future. This is typically accomplished through two methods, information on current behavior and goal progress information. We argue that many of these tools and services are not taking full advantage of the vast amount of information that is available to them.
The wide-spread proliferation of application programming interfaces (APIs) that allow developers and users to access large amount of data opens up numerous possibilities for possibly improving the health and behavior conversation between a user and his or her tools/system of choice. We foresee unique opportunities to use historical behavioral data, contextual information (e.g. location, social interactions), and health actions to highlight patterns and provide feedback through three mechanisms: 1) reminders of success, 2) behavioral prompting, and 3) contextual reminders.
The road to good health is not an easy one and there are numerous examples of individuals who unfortunately lapse into negative or poor behavior patterns. We are proposing that when “failure” points are identified there is an fantastic opportunity to remind the user of previous success. Reminding a user that they have had success in the past may help to limit self-doubt and reductions of self-efficacy. The psychological burden associated with failing to meet goals could be quickly replaced with a positive a reminder of the user’s mental and physical capability that is based on objective historical information. Instead of just having an empty “You can do it!” we envision future services that say, “We believe you can do it because, look, you’ve done it before!”
We also see the potential for building upon the concept of modeling illustrated in social learning theory and social cognitive theory. While modeling is typically thought of in the social sense, we propose that services can use historical data and contextual information to create powerful and meaningful representations of a user (maybe as a digital avatar). By presenting a user with their past self they can use it as a tool for comparison (“What am I typically like?”) or competition (“How can I be better than my previous self”). Imagine, for example, waking up in the morning and seeing your past self and associated behavioral data in your bathroom mirror or on a display on your refrigerator. We believe that this past self could act a positive guide to help you lead a healthier life.
Lastly, the large amount of information stored in your behavioral databases has an inherent ability to converge and provide information about contextual factors associated with behavior. For example, we can easily find out if you get more or less steps on days it is raining or if you tend to eat worse when you check in to airports around dinner time. Using simple data mining and contextual linking it is possible to identify positive behaviors patterns and bring them to light. By tapping into the rich digital histories being captured and stored across many services we may not only help a user remember, but also enhance their ability to celebrate and re-enjoy healthy behaviors.
Too often, we encounter warnings of services tracking out behavior and using if for their own personal gain. It is time that we ask the tools and applications we use to help us lead healthier lives by taking full advantage of the vast amount of historical information we are collecting. The Spanish philosopher, George Santayana told us, “Those who do not remember the past are doomed to repeat it.” Our increasing digital lives allow use to not only remember the past, but harness that powerful information to help us lead better, healthier lives.
This article is a summary of a position paper by Ernesto Ramirez and Eric Hekler that will be discussed at the Personal Informatics in Practice workshop at CHI 2012 in Austin, TX on May 6, 2012. The workshop will be a gathering of researchers, designers, and practitioners exploring how to better support personal informatics in people’s everyday lives.