One reason I wanted to test various health tools was to get a better handle on my sleep. People are pretty much in agreement that sleep can influence your productivity and error rates. But, what constitutes good sleep and can you score it? Below is a quick rundown of how each of the 4 devices I’m testing displays my sleep phases and ranks my sleep.
Quantifying Sleep Scores & Caveats
Chances are you’ve probably heard the phrase, “numbers don’t lie”. I’m not going to argue that point, but throw out corollary #1 – “they don’t always tell the whole story”. The example night below shows each of the 4 devices I’m testing and how they quantify the same night’s sleep and provide an overall score. These numbers have more value when used to compare sleep over time using the same device and methodology. In other words, a MyZeo ZQ score of 85 shouldn’t be compared to a WakeMate score of 67.
Now that I’ve gotten the disclaimer out of the way, I’ve got to say I like seeing an overall sleep or efficiency score. I tend to glance at the number and then dive further into the stats if available. Currently, the only device that immediately shows this score is the MyZeo clock when I dock my headband. The other devices require that I sync the data and go online.
Sleep Stages and Patterns
As you might expect, the MyZeo and WakeMate provide the most amount of sleep data because they’re designed solely to analyze your sleep. In contrast, the BodyMedia FIT armband and FitBit Ultra take a more holistic approach and sleep is just one dimension. I also suspect that these devices capture a lot more data than they display, probably for ease of use.
(Note: The full report that comes out in December will cover all these devices in more detail including additional sleep reports.)
For me, the Zeo Sleep Manager provides the best sleep breakdown. I can’t say the picture above will reveal the secrets of how to sleep better, but it’s easier for me to correlate the colored bars to what may have occurred during the night or some other factor. For example, I know that the double orange bar around 4:30AM maps to the time I heard a noisy neighborhood cat. The daily analytics also makes me question what was different about that night because I recorded much more “Deep Sleep” than normal.
Like the MyZeo, the WakeMate is designed for sleep analytics. The reporting isn’t as detailed as the MyZeo, but it does quickly compare various metrics to my lifetime scores and others. Not sure who comprises this global community and how many people are in it.
The FitBit Ultra gave an entirely different picture of that night’s sleep that makes me issue Corollary #2. “First looks can be deceiving”. If I had to go on FitBit’s sleep analytics by themselves, I might worry when seeing 54% efficiency and an actual sleep time less than four hours. As it turns out, there are 2 sleep settings and my device was set to “sensitive”. This meant the device was more attune to my tossing and turning. I’ve changed the setting and suspect things will normalize. This also highlights why I’m testing for 2 months before writing the report.
While the MyZeo gives you a detailed view of your sleep, the folks at BodyMedia take more of a 30,000 ft. approach to sleep. Sleep is just one facet of your health along with exercise and nutrition. The graph quickly answers the question of whether I met my sleep goal. The goal is defined by my preferred bedtime and rise time.
One aspect I find appealing about the BodyFit Media device is that I don’t have to do anything special other than wearing the armband. For example, I tend to take naps and like to read in bed. With this device, I don’t have to press any buttons to distinguish my activity or control a timer. It can figure it out on its own.
Finding the Ultimate Sleep Algorithm
While it would be nice to have one sleep algorithm that works across all devices and users, it’s not here so dream on. I’m not even sure the scientific community has agreed on a standard measurement, but it seems most of these scores put a fair degree of weight on the amount of time in bed.
Last Updated (Saturday, 14 January 2012 18:01)