Quantified Awesome: Data from waking up 3 hours earlier than normal


I’ve been taking advantage of jetlag to wake up earlier than usual. I go to bed when I’m tired and wake up when I’m ready. Most of the time, I wake up on my own, although I set my alarm for 6 AM as a safeguard. From 2011-10-15 to 2011-10-28, I woke at 5:41 on average, with a median of 5:12 and a standard deviation of 1:09. (Weekends…)

I liked being up early in the morning. No guilt about hitting snooze, no rush to the bathroom, some time for personal projects or work momentum before the workday starts.

I was curious about a few things:

  • Around what time should I plan to be in bed if I want to continue waking up early?
  • Does waking up early actually give me more discretionary or work time, or do I give up time because I get sleepier in the evening?
  • Waking up early usually means I’m tired at the end of the day. What kind of an effect does that have on the things I feel like doing?

Fortunately, I’ve been collecting time data for a while, so I can compare it with a similar two-week period where I’d wake up around 7. Let’s see how the data stacks up against the data from 2011-08-20 to 2011-09-02, a two-week period also without any long weekends.

Going to bed

n = 14 for each condition

Late Early
Wake-up average 8:33 AM 5:41 AM
Wake-up median 8:30 AM 5:12 AM
Wake-up stdev 0:29 1:09
Bedtime average 1:03 AM 10:16 PM
Bedtime median 0:59 AM 10:09 PM
Bedtime stdev 0:38 1:15
Sleep average including naps 7.8 hours 8.2 hours

So if I want to get up at around 5:30, I should be in bed by around 10.

I tend to sleep less when I sleep later. Work gives me a reason to pay attention to my alarm clock, so even if I hit snooze and have lower-quality sleep, I’m out of bed. I used to stay up late so that I could spend more time hanging out with W- or working on projects. Sometimes it took me a while to go to bed because I’d get carried away hacking. My standard deviations for the late condition are pretty low – mornings because of the alarm clock, and evenings because I eventually look at the clock and go "I really should be in bed."

Waking up early and going to bed when I’m reasonably tired means wasting less sleep time fidgeting, enjoying better sleep and morning quality, and less stress in the morning. Based on this two-week sample, the difference is around 24 minutes a day. It’s hard to tell whether the increased wakefulness of being up early and on my own schedule compares well with the stress and bleariness of mornings jumpstarted by an alarm clock and stress. I think it’s worth it. Besides, if I zoom out and look at more than just the two-week period – say, two-month period 2011-08-01 to 2011-09-30 – I find that I actually sleep around 8.0 hours on average, so the difference isn’t that great.

I wonder if jet lag affected my numbers, increasing the amount of time I needed to sleep. If I look at just last week’s data (2011-10-22 to 2011-10-28), though, it turns out that I ended up sleeping a little more after I had theoretically recovered from jet lag: 8.7 hours on average, or 8.2 hours on weekdays and 10.1 hours on weekends. Part of that might be due to the Halloween event we photographed on Saturday – I got a case of introvert overwhelm and napped for 4.5 hours afterwards to recharge. No significant differences, though: unpaired t-test between number of hours slept during first week (M=7.6, SD=1.5) and second week (M=8.6, SD=1.4); t(12)=1.49 p=0.16. We’ll see how the numbers work out as my routine stabilizes.

I do like the sleep quality. People can spend a lot of time and money in the quest to improve their sleep. I had been playing around with using eyemasks or eye pillows I’ve made myself, and had even considered getting a light-based alarm clock and/or blackout curtains. Going to bed when I’m tired means not needing any of those things, so I can save that money for other things.

In a previous experiment with early-morning wakeups, my husband and I noticed that our schedules were diverging a bit. He’d stay up late, I’d wake up early, and we had less conversation time. He was similarly jetlagged this time around, so we’ve settled into a good routine with plenty of time in the evening and some high-quality morning time too. It’s been working well.

Waking up early and discretionary time

Late Early
Discretionary time average 6.4 hours 5.7 hours
Weekday average 5.2 hours 3.8 hours
Weekend average 9.4 hours 10.6 hours
Work 5.7 hours 6.1 hours
Weekday average 8.0 hours 8.4 hours
Weekend average   0.4 hours
Unpaid work average 1.9 hours 1.2 hours
Personal care average 2.4 hours 2.8 hours
Sleep average 7.8 hours 8.2 hours
Discretionary + work 12.1 hours 11.8 hours

For productive time, I looked at the sum of time I spent on work and the time I spent on discretionary projects. This took into account the extra time I shifted towards working last week. It turns out that there’s a little difference between the discretionary + work time I had (late: M=12.2, SD=2.0; early: M=11.8, SD=2.1), but it’s not significant either (unpaired t-test t(26) = 0.45, p = 0.65). So it looks like waking up earlier doesn’t mean giving up too much – or gaining a lot – in terms of focused time.

Activity substitution

If I have about the same amount of discretionary time anyway, does waking up earlier affect the kind of things I spend my time on? This one is harder to figure out, because other variables affect how I spend my time. I spend more time drawing when I’m attending events or preparing for presentations. I spend more time working on my personal dashboard when I’m buzzing with ideas. I spend more time writing when I don’t have lots of posts queued up. It’s hard to say.

Looking at my time graphs, though, I see that when I woke up early, I didn’t really have the chunks of discretionary time that I’d hope to have in the mornings, and my evenings were more fragmented and other-focused. When I woke up late, I tended to have more me-time at the end of the day, and I still had enough energy to make the most of it.

A previous analysis showed that even when I stayed up late, I didn’t really have many discretionary activities that used a four-hour chunk of time, so waking up early doesn’t mean I’m missing out on activities that need a long chunk of time. However, after-school hours tend to involve discretionary social activities, and I usually carve out time for personal projects either late at night or for a short time in the morning.

Overall, I’m happy with how I spend my discretionary time. I feel like I’ve made reasonable progress on my personal projects, and I’m glad I’ve been able to help with things like homework. I might shift things around so that I can write and program more, probably when I get work back under control. As we improve routine processes like cooking, we’ll free up more time for other pursuits too.


It looks like waking up early doesn’t have a significant impact on how much time I sleep or how much focused time I have. I like the sleep quality and the lack of stress in the mornings. It might come at the cost of not having a longer window of discretionary time focused on personal projects, but social time is good too. Overall, I’m happy with waking up early and the resulting shifts in my schedule, and will continue waking up early and going to bed late.

Testing your assumptions and trying new things is much easier when you collect data. I’m thinking of sharing observations every Monday. Check back next week for more!

Photo of Toronto at dawn © 2009 Mac McGillivray, Creative Commons Attribution License

One Pingback/Trackback

  • Interesting analysis. This is what I want to do as soon as I have enough data points. For now, just collecting data, I still need to write some code to generate the duration column… And then some plots (I want to see a mosaic plot of my activities).

    By the way, the interactive graphs don’t work :(


  • Ruben: Right! Let me fix that. Should work now.

  • May

    Hi Sacha, Greetings from singapore!!

    Wow!! I love your positive attitude and way of living! I love your blog name ” Living An Awesome life!! SOmetimes Life is not smooth sailing afterall, but these imperfections make us appreciate life and how awesome it can be actually! Its like you never know how Sweetness tastes like until you tasted Bitterness/Sourness. I have never seen stats being linked to sleeping hours before . Although Im trying hard to figure out some of the stats still, nevertheless your post amazed me! Looking forward to more posts coming right up!


  • Pingback: Quantified Awesome: Analyzing time data–the questions I ask and how I answer them » sacha chua :: living an awesome life()