What is {meta}-Data+Life

Like many people in the US today, I have a tracking device on me nearly all the time. The little watch-like computer strapped to my wrist captures a lot of data about me.

The other day I was talking with someone and the subject of ‘steps’ came up. They were trying to make sure they got in their steps for the day. They commented about how my dog walking probably keeps my steps up.

Most of us have a conversation similar to this at least once a week, but it got me wondering: What is a ‘step’ anyway? Why on earth does this measurement mean anything to us?

As a data person, I build processes that make data useful. There is a lot that goes into this. Ask a data person what they do and prepare yourself for an onslaught of really nerdy technical jargon that will glaze over your eyes … unless you are also a data person.

Deep in the heart of any data person is a passion for good measurements and good metrics. (No they are not the same.)

How are ‘steps’ a good measurement?

It is less important that a measurement is ideal, or even good, than our ability to make it useful. Any information you collect is just data floating around in your skull until you have a use for it.

We faithfully track our steps because that is what the machine on our wrist is capable of measuring. There are other things required before that measurement is useful to build a metric, but the makers of your device and app thought of that already.

If you dig in to the app on your phone that ties to the fitness tracker on your wrist you will find lots of things about your steps. Mine tells me my average step length. Well now, how do they know that?

Here is where I start guessing based on my knowledge of how data works together to create meaningful metrics (now we are moving past raw measurements).

To know average step length, they connect the GPS distance traveled to the raw number of steps, and divide. Boom. Average step length. Easy, right?

So what is the point of this data?

Metadata is data about data, which is why it’s meta. It’s not the measurement itself, but information that relates to the measurement. I think of it as the context of data.

My watch tells me how many calories I burned, and that is something I, or any consumer, can associate easily with food intake (calories being metadata about food). Getting to calories burned requires even more information to connect the steps to something like a ‘work effort of each step.’

Data like your age, height, and weight are tied through formulas to the steps. I’m sure everyone put that into their phone when it asked them for it, right?

The formulas for calorie burning are based on the type of activity you are doing (some of these devices have to guess somewhat because you are moving fast, or not moving at all).

What comes out of all this imperfect data is something that is more useful than just the count of steps you took. Step count is just a single data point.

Very little, if any, truth comes from a single data point.

It is said that when something happens one time is an event. When it happens several times it looks like a pattern. When it happens a lot of times, reliably over time, it is a trend.

A step, by itself, is nothing. Collect enough of them over enough days, along with a bunch of other measurements then you can build a metric. Will it be “right” or “accurate?” Maybe not.

It might be useful.

I’m sure if you look around you will find a lot of inaccurate, incomplete, or seemingly detached (random-looking) events in your daily life. Taken alone, they might be nothing.

When you connect the measurements, imperfect and incomplete as they may be, you can make them useful for a purpose.

The purpose, or the useful goal of our effort, is what counts in the end. Meaning is the beginning of every endeavor. Every measurement is only as good as its utility.