The Limits of Democratizing Data Collection

There is an interesting article by Gary Wolfe at NYTimes (link) about people’s self-tracking behavior and the data that is accumulated. People count calories, document fitness regimens, keep calendars, and on and on and on. People have long tracked their behavior but now, rather than using pencil and paper, it is digital, often widely available (e.g. through Facebook, Twitter, 4square, MyFitnessPal), and it’s in a form that can be analyzed. This last piece is crucial, while the accumulated data is in a form that can be analyzed, some people might not have the knowledge to do so — a point made in the piece and the focus of this post.

Analyzing one’s caloric intake is probably not that difficult, but new technologies (e.g. Zeo) are allowing people to keep track of other behaviors, like their sleep/wake cycle via electrical scalp potentials (link) or through the accelerometer in their iPhone (link). Part of the Zeo sales pitch is that a person can wear a device that measures brain activity which can then be translated into their own sleep/wake cycle. From this, the user can determine, and get coaching, about how they can improve the quality and amount of sleep they get.

I’m immediately skeptical of collecting brain activity from one site as is the case with the Zeo, but this is beside the point. My point is how does a person who might have little or no knowledge of sleep/wake cycles interpret this data? Zeo may have skirted this issue with “coaching” but the iPhone users are on their own to interpret the data.

And this isn’t just a problem for lay people trying to analyze personal data, experts might run into a similar problem — data without enough understanding to interpret the results. The podcast Radiolab recently interviewed two PhD students about a computer ‘Eureka’ that they developed to sift through large amounts of observations and find laws that govern, or explain the data. Another researcher, Gurol Suel, used this program to analyze observations from a particular bacteria in order to determine the relationships between the nutrients, proteins, and other components of the bacteria. The equations that resulted described the data and could predict what the bacteria would do, but the problem was, and still is, interpretation — WHY does the equation work the way it does. fMRI data has been similarly criticized. In fMRI’s early days, researchers and news outlets would place two brains side by side and explain that one was ‘normal’ and the other was after X disease/drug/activity. The question then is “So what? The two brains are different, but what does it mean?” The situation is improving for fMRI data because we know more about the brain then ever before. But what about ‘Eureka’? The computer might produce brilliant models of behavior or biology but without the understanding to test and ultimately understand the equations, is it even worth having them?

Normal Brain (left), and Cocaine Brain (right).

This is probably a natural scenario for scientists who are using new technologies empirically. With insight and hard-work our understanding will catch up with the technology. Lay people, on the other hand, who attempt to draw conclusions about their personal data will likely not put in the work or have the necessary insights to ever understand the information they collect about themselves, after all it’s not their job to do so.

-Posted by Tyler

  1. “Lay people, on the other hand, who attempt to draw conclusions about their personal data will likely not put in the work or have the necessary insights to ever understand the information they collect about themselves, after all it’s not their job to do so.”

    Objectivity + critical thinking + data = ?

    Wonder what is on the other side of this equation.

  2. Objectivity + critical thinking + data = Conclusions that can generalize to a larger population.

    One redeeming factor about the data that lay people collect is that they might also simultaneously collect subjective reports about their caloric intake or sleep patterns. For example, the iPhone sleep cycle application user could also annotate the figures the application creates with information about their daytime sleepiness or time to fall asleep at night. If they did this, they could make very relevant conclusions about themselves.

    And maybe this is the real message, if done correctly (i.e. with annotations) lay people can collect data and make relevant conclusions about themselves. Whereas, experts collect data and with any luck make conclusions about the population.

  3. What’s Taking place i am new to this, I stumbled upon this I’ve found It absolutely helpful and it has aided me out loads. I hope to give a contribution & help other customers like its aided me. Great job.

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: