In late August, I bought a Fitbit One. One feature I was curious about (but wasn’t the reason I bought it for) was sleep tracking.  Was it accurate or useful?  I wasn’t sure, but it was easy to use, so I thought I’d give it a try and see. To activate, just before you’re ready to go sleep for the night, press the button on the Fitbit until you see the stopwatch start (a recent firmware update results in a brief buzz feeling when you activate the stopwatch.  I imagine that’s the same thing it uses to awaken you in the morning, but I don’t use an alarm, so I don’t know, but that was a very welcome addition to the Fitbit’s repertoire). When  you awaken in the morning, press the button again until the stopwatch stops (you’ll feel a buzz on this latest firmware), and that’s it.  Next time you sync the device, it uploads the information to your dashboard or your phone app, and you have your sleep data.  If you pay for the Premium service, you can download this information into a spreadsheet.  I’m cheap and hand entered the data I wanted.

I have data from August 23, to October 24 in that spreadsheet.  I have a few comments, and I’m ready to make a few observations about it.

Comment 1:  The Fitbit One includes a wrist strap to wear at night.  It works fine, but I thought it was a pain to deal with and after a few weeks, I just started clipping it to the neckband of my pjs. When I did that, I noticed a decrease by about half in times awakened being recorded.  I switched the device to “Sensitive” mode for a couple of nights, but that didn’t change the number of times awakened, but my Sleep Efficiency plummeted from the 90th percentile to the 70th percentile, and I didn’t perceive that my sleep was any less efficient, so I changed the setting back to “Normal.”

Comment 2: By and large, the data recorded is fairly consistent with what I perceive my sleep quality to be.  Your mileage, of course, may vary.

Here are the pieces of data I’m recording on my spreadsheet and what I’ve noticed about them over the last two months:

Sleep Efficiency (85% is considered normal): As I mentioned earlier, my percentages are usually in the 90s.  I’ve had a few days in the 80s and about an equal number in the 70s.  In one particular instance, I had a truly awful night’s sleep.  I woke up at 1:30 am and couldn’t get back to sleep for two and a half hours.  That showed in the graph, and I firmly believed the 72% sleep efficiency recorded for that night.

Time to Bed: This is when I turned on the Fitbit sleep tracker.  It varies from 7:30 pm (I was sick) to about 3 am (I was up reading).  My goal is to get six hours of sleep per night or more.  I’ve done that for all but 11 nights of the last two months, and only four of those nights have netted me less than five hours of sleep. Time to bed doesn’t seem to make much difference in quality of sleep.

Sleep Latency (Time to Fall Asleep.  Normal is 15 minutes): This is really fascinating information.  Only twice have I taken longer than an hour to fall asleep. It’s recorded as little as one minute!  Surprisingly often, it takes me less than five minutes to fall asleep. Fourteen times over the last two months, it’s taken longer than 20 minutes for me to fall asleep.

Times Awakened: With the Fitbit on my wrist, I averaged about 20 times a night.  With it on my neckband, it averages around 10.  One night last week, it recorded a zero! I went to bed very late that night and I slept well. That’s not a typical night.

Time in Bed: I don’t really do anything with this data.

Actual Sleep Time: As I mentioned above, I strive for six hours or more of actual sleep time per night.  I’m generally pretty good at achieving it — unless I stay up reading until 3 am.

Notes: This is where I note anything unusual about  the night — I had to get up at 3 am to turn on the soaker hoses in San Antonio.  I was sick.  I did something unusual with the tracker.  Whatever I think I need to note about the night or some stat from the night.

Overall Assessment:  I bought the Fitbit to replace my aging Omron pedometer.  I wanted a pedometer that would automatically synchronize with an app to track the data. Of course, the Fitbit does this very well.  So do a number of other similar products.  I didn’t conduct extensive research. Admittedly, someone posted about theirs and several other people chimed in about theirs, and I was swayed to research the Fitbit.  The next day, we went to Best Buy, and I bought the one I’d chosen.  I’ve been happy with it.  It lasts about a week to a week and a half on a charge.  The sleep tracking was an unexpected benefit that I’ve found very interesting.

For an added bit of fun, some people have been hanging extra Fitbits on their cat’s collars and Tweeting the Fitbit results for their cats.  I linked my results to my fictional polar bear account on Twitter, so if you’re bizarrely curious about such things, tune into the @InuitTheBear account on Twitter.

Fitbit makes four products:

Zip ($60): replaceable battery, less expensive, doesn’t track sleep, and is more like a true pedometer but does sync data with the website and smartphones

One ($100): The one I use with a rechargeable battery, display, fits in pocket or clips on an item of clothing, tracks sleep, syncs with website and smartphones

Flex ($100): You can wear it on a wristband like a watch, but it only has LED lights for a display to tell you where you’re at in achieving your daily step goal.  Very minimalist, but it does everything the One does.

Force ($130): A new product just released which is basically a One and a Flex combined.  Or, a Flex with a display.  Pretty cool.

According to some people who have tested both, they think a One carried in a pocket or clipped to clothing at the core of the body is a little more accurate than a Flex on a wrist.  For that reason and wanting a display I could read, I chose the One over the Flex, and would probably choose the One over the Force after I’d considered those details. I have to be careful not to put anything else in my pocket with it, because the face scratches easily. All said, I love the Fitbit One (it’s about one-third the side of my Omron pedometer).