I typically go to bed around 10 PM and leave the bed (for the last time) around 7 AM. This is the duration of time I've been using to calculated TIB (Time in Bed).
My typical sleep is punctuated with 3 pee breaks, the last of which is usually around 530 AM. By that time in the morning I'm usually fully rested but since it's still early in my work schedule, my habit has been to stay in bed and follow my breath (in pseudo meditation) or slip into a kind of hynogogic reverie, until around 7. So my TIB INCLUDES this last hour of "wakeful rest" even though I know I am for the most part awake only lying prone in bed instead of walking about or seated.
To my thinking this last hour should not affect the apps' calculation of my total minutes in REM or DEEP sleep. Nonetheless, it does so, in a dramatic way.
I've attached 2 sets of images to make my point. These show how the app calculated today's stages of sleep. (This pattern of discrepancies can be demonstrated for any other day's data too.)
The 1st set of images shows TIB until 7:15 AM. Time Asleep is given as 7 hr 59 min. The calculations in this first set of data INCLUDED an hour and a half of meditation/reverie I did this morning while in bed. Notably, REM time is given as 44 minutes, DEEP sleep as 65 minutes. If you look at the image, you'll see there were ZERO spikes in the graph showing DEEP sleep during that last hour and a half. Also, the total REM shown is about 8 minutes worth for the hour and a half.
For the 2nd set of images I edited out the final hour and a half of TIB during which I was mostly awake.
The amount of REM sleep calculated for the 2nd set is dramatically different. Now I supposedly had 83 minutes of REM which is almost double the 44 minutes when I included MORE TIB. That's absurd. Given less data and less time for sleep how can I have double as much REM time? If you compare the 2 sets of images for the stages it makes no sense for there to have been double the REM time.
Also, the minutes in DEEP sleep seem to have changed inversely to the REM numbers. While the app showed double the REM minutes (83 v. 44) when I set the wake up time to 5:48 instead of 7:15, the DEEP sleep time moved in the opposite direction (from 65 minutes down to 42).
Why are there such discrepancies in the apps calculations based on length of TIB when during the last hour and a half in bed the images showed nearly no REM or DEEP activity? So if one removed that period with almost no REM/DEEP activity why would the recalculated data be so different from the full TIB calculation?
Best Answer
When I wake up, I copy, y sleep data into the Sleepmeter app to make calculations, so I prune off any excess Awake time at either end. It often moves the sleep stages around after that, even though all I've done it set it to begin and end when I'm actually asleep. It's some sort of bug. Fitbits are very approximate at calculating sleep stages, anyway. There's been research showing that they're way out for deep sleep in particular. They don't have electrodes on your scalp, they can't actually calculate REM and such.
If you're dozing for the last hour, that's stage 1 sleep. People often think they're awake when they're not during that type of sleep. I get this in particular with naps, and oftrn start my naps by meditation. It's interesting to see which times Fitbit thinks I was napping, and which times I'm convinced that I never quite got to sleep. I'm taking it as rest time either way.
Plugging the data into Sleepmeter is fun, by the way, since you seem to be a fellow graph addict. Here's the overall chart for my last three months, although I have way more graphs than that, including ones to show me how different factors influence my sleep.
Best Answer
Thanks for the heads up re: Sleepmeter. I created my own tracker a while back which uses Excel's standard deviation calcs to alert me when a night's results are way off from my averages.
Best Answer