Random (but not really)

Wednesday, March 25, 2015

Pedometers: Data Comparison!

Another quick reminder of what I’ve been testing:

Garmin Vivofit
FitBit One
Omron HJ-323U

The Devices
The Software


I tracked the output of the various devices for three weeks, noting steps, kcal, and miles at various points of the day, including immediately before and after exercise or hiking.

Here’s what I found.



As I’d found previously, the FitBit overcounts steps. Accupedo and the Omron are generally pretty close to each other, and typically had lower step counts than the VivoFit and the One.


On days when I was doing things like cleaning house or baking, the VivoFit had a much higher count than any other device. (I also threw out the Accupedo data on several of these days, because I didn’t have the phone in my pocket for several hours.)

I’m not terribly distressed that the VivoFit had a higher step count on days when I was baking or cleaning, because I’m active during those times, just not moving from one place to another.

Calories Burned


Not only does the FitBit overcount steps, but it also generally overcounted how many calories I burned during the day.


Here you can see that during periods of exercise, the FitBit gave me a far higher calorie burn than any other device–including the Garmin Fit, which should also have been taking elevation and pace into account. (Note that for walks, the Omron and Garmin Fit had very close calorie burns, but for hikes (where we were in steeper terrain) the Garmin Fit gave a much higher calorie burn than the Omron. I believe the Garmin Fit data should which should be fairly correct, although I know that using the heart rate monitor would make it more accurate. But I find it uncomfortable, so you don’t get that data because I am unwilling to suffer that much for data.)

Putting that together, we can see how calorie burn and steps tracked for my lunchtime walks.


The step counts are all relatively close during these time periods, but the FitBit in general gave me a much greater calorie burn (I’m not sure what happened on that final Thursday–I’m guessing operator error in noting the counts.)

Yes, I do walk up and down stairs during my lunch walks, but I don’t think that’s enough to account for the FitBit calorie burn being that much higher.


And finally, because I can’t help myself, I compared the mileage calculated by the pedometers as opposed to the distance given by the GPS for outdoor walks and hikes.


GPS readings were from the Garmin Fit, which tracks mileage using my phone, and my Garmin Oregon GPS. I’ll note that the Oregon GPS does tend to give me a slightly shorter reading than is probably true, simply because I meander. So fractally, the distance I walk is probably longer than the distance calculated by the GPS, but not by a large amount.

Also interesting, the Vivofit seems to underestimate distance when I’m hiking and walking, as compared to other devices, but not consistently so, which could be due both to the terrain, and to how frequently I stop to take pictures when I hike.

And to be fair, hiking is quite different from walking, since one’s stride is generally not consistent.

So there you have it: The FitBit continues to overcount steps and overestimate calories burned, but the interface and community are so much nicer than any other device, it’s still my current favorite.

If you do outdoors activity, the Garmin Fit app is very nice, and I can definitely recommend it.

And if you don’t want to carry around an additional device, the Accupedo app is a very good alternative to a separate pedometer (assuming you can correctly calculate the sensitivity).

Previous Pedometer Geekiness:

Powered by WordPress

books main pictures cats e-mail