Brett Rants

Sleep Studies

Using Absolutely Horrible Data.


A bar graph showing when I was sleeping for a few weeks, gray is fuzzy time, sort of laying in bed, black is sleeping, it is truly horrible data, data collection when one is on the edge of sleep is not the easiest thing

It's been well over a year since I collected this data. Sort of ironically, a loved one was talking about procrastination and this project rapidly came to mind, as I've been sitting on it (like) forever. But I suddenly found the will right after our conversation and pounded out a graph. That little thing of beauty (as I will be kind to myself, even if you will not) took me 1-2 hours. And this was after I'd already prepared the raw data and generated the following preliminary analysis.

Sleep Studies Summary

DAYS: 23.29


TOTAL SLEEP: 199.15 Hours

Please see the raw data (a link is available at the top of the page) for further elaboration. But then, the only item that should be confusing is the FUZZ (as shown in gray in the graph above). Well, that's what I was calling that laying around in bed time, when one is not quite asleep, but trying to be so. It's not an accurate reading in the least, as I often lay awake in the middle of the night for 1-2 hours. So, like, I know the data is horrible.

I determined the Average Sleep (per day) by dividing the Total Sleep by the number of Days. But trying to determine how much I slept in a typical night was more troublesome, because I had to back out naps and those times when I was awake in the middle of the night.

To try and figure this out, I eliminated shorter spans of sleep. For instance, if I removed all periods of rest less than an hour in length, I wound up with 31 sleep periods, averaging 6.2215 hours each. And with that explanation, the other possible ways of interpreting the data should make more sense.

NO NAPS(<1)=(31): 6.2215 (ave length)
NO NAPS(<2)=(26): 7.1673 (ave length)
NO NAPS(<3)=(24): 7.5625 (ave length)
NO NAPS(<4)=(23): 7.7384 (ave length)
NO NAPS(<5)=(21): 8.0714 (ave length)

After this long, I can't say I care about the data, the analysis, or anything else related to this project in the least. Further, I question whether I learned anything... except for maybe how hard it is to manually collect sleep data on oneself and a future personal preference to collect data (any sort of data) in a full-on spreadsheet rather than as an iPhone note. But then, the iPhone is by my bedside, while a computer is not. And as such, I still find my iPhone (as opposed to my Android... for legacy reasons, I am sure) to be my preferred form of dynamic (so, like, in the wild and on the fly) mode of data entry.

next Brett Rants entry

Home Brett Rants Index

Hey! Maybe (someday), I'll do a write-up on my movie watching habits. But as of this writing, that dataset is a total train wreck. And I'm even using a spreadsheet for that, so go figure.

© copyright 2019 Brett Paufler