The Limits of Data

I’m a big fan of data.  I read The Circle, which kind of creeped me out, but I also loved the access to data the book explored.  I often find that when I look at data, I see things I didn’t expect to see.  For example, I look at tests and how people answer questions.  The questions I think are hard sometimes have the most right answers.  It doesn’t mean the question isn’t hard; it means there’s another story there.

The data issue from my last post is a case in point.  That’s raw data there, but it’s been interpreted by someone (or group of someones) in a certain way.  Subjects and industries have been grouped arbitrarily.

Often people bring up data as a way to solve problems.  If we just have enough data, then we can fix X problem.  Think about data and students and schools.  We’ve been trying to use data to create better schools forever.  It’s hit or miss at best.  Data only tells us so much.  And what it does tell us is often our own interpretation.  It’s rare that you crunch some numbers and a graph appears and it’s suddenly clear what needs to be done.

I’ve been thinking about my own data lately as I try to lose the handful of pounds I’ve gained in the last few months.  I invested in a fitbit to help.  And I’ve gone back to tracking my calorie intake and my weight.   What my initial data tells me is that I’m pretty sedentary naturally, that I have to make myself move.  I don’t think that’s a desire on my part, just a factor of where and how I live.  I have to drive most places to get stuff.  To walk more than a couple of miles, I have to plan it.  It doesn’t happen naturally.  Also, I eat more than one might think.  I like food and cutting back is a challenge.

Looking at losing weight as just a numbers game can be helpful. It’s useful to know that I went 162 calories over today based on my intake and my activity.  But knowing that doesn’t necessarily make me want to get out of bed right now and go for a 2-mile walk (which is what I’d have to do to burn off those extra calories).   There’s a human element to all of this.  Yes, the data gives me valuable feedback, but I have to do something with that feedback.

Data can tell us all kinds of things, but it’s still up to us to either interpret the data or act on it or both.  And sometimes our interpretations or actions are wrong.  And then what?