Sleep Experiment - Day 7

June 27, 2011 – Day 7

Today was pretty great. In fact, it felt like clockwork. Every nap was preceded by a brief period of tiredness and followed my only a short time feeling groggy. It seems that yesterday was indeed just a random blip in the adjustment process.

I once again faced a few issues regarding sleep scheduling in the face of a more active schedule. I had to take my first nap earlier than usual and my second nap later than usual. This resulted in more extended periods of tiredness after and before these naps respectively. However, it seemed that my body was able to readily adapt to these minor changes.

As opposed to yesterday, I experienced nothing but the positive effects of polyphasic sleep: I was well rested, I felt alert, and I was able to actually use all of my waking hours for productive endeavors (I didn’t really take advantage of this though). Riding this wave of seemingly positive results I decided to break my own rule and start analyzing my data from the PVT. Things haven’t been going exactly as I expected. But more on that later.

The breakdown of my day is as follows:

6am – wake up. I woke up quickly from my alarm. I felt the usual grogginess but this time it passed almost immediately following my walk (longer and more rigorous than usual) and breakfast. I also had course registration to deal with this morning, which may have contributed to the alertness.

10:45am – 1^st nap. Usually my naps are at 11am but I was invited out to brunch and had to push this one forward. I wasn’t particularly tired when I got to bed but I still managed to fall completely asleep.

11:05am – wake up. I found it especially easy to wake up from this nap. I had to rush out to brunch and I was afraid of oversleeping. It took me a little extra time to fully wake up (especially because I did not eat soon after waking up) but I was soon at my normal state of wakefulness.

4:30pm – 2^nd nap. I began to feel tired around 2pm this afternoon. It wasn’t a terrible feeling and didn’t hugely impact my afternoon but it was a little annoying. I also delayed my nap by 30 minutes and was thusly awake for an extra 25 minutes or so between naps that I don’t usually have. I managed to fall asleep very quickly.

4:50pm – wake up. I woke up quite quickly and felt almost immediately well rested. I had a very productive evening following this nap.

9:40pm – 3^rd nap. It took me a while to fall asleep during this nap. I eventually fell into a deep sleep.

10pm – wake up. While I woke up with my alarm, it took me a few extra minutes to get out of bed. I have felt a little groggy since waking up but that is due to the sedentary activities I have done this evening.

The data for today:

While I felt fairly alert all day, the PVT seemed to be disproportionately difficult. I felt that my reaction times did not necessarily reflect my wakefulness. I have also decided to scrap the working memory portion of this experiment. I don’t believe that I would have gained any useful data from performing these tests without a baseline and some method of scaling for different n-values in the n-back task.

While on the topic of data, I finally did the analysis for the last seven days of data (the days of polyphasic sleep). I’m not entirely sure how to describe what I saw and am nowhere near a point at which I can make any definitive claims. But, I can give the general gist of what I’m examining and what I’m seeing.

In completing the PVT each day I only get one piece of data: an approximately 20 item table of reaction times. To simplify things and to allow for comparisons across days the sample mean and variance is calculated (as an unbiased estimator) for each days data set. I’ve also been calculating the median of the data (although I don’t know what I’ll use this for, if anything).

The most basic analysis that I have done is plotting the mean reaction times (with error bars) versus the day of the experiment. This plot (I’ll include it in a later post) looks like a scatter about a constant value. I haven’t experimented with any fitting routines as of yet but this plot doesn’t really show anything exciting.

I have also been rating my sleepiness on a scale from one to ten. While this is extremely arbitrary (I found some rubrics that actually rate sleepiness, but again this was a few days into the experiment) it still gives a general idea as to my level of perceived wakefulness.

Plotting this ‘sleepiness ranking’ versus the day of the experiment has not been very illuminating either. Even when compared to the reaction times. More detailed analysis will need to be done to potentially show a correlation between perceived sleepiness and reaction time.

The final item I am collecting is ‘lapses in attention’. This is an item often seen in the articles I have been reading and is easily extracted from the data. I simply define any reaction time over 0.5s to be a lapse in attention. This value may seem arbitrary but it is the value that seems to separate the max of my reaction times on well rested days (ranked as a four on sleepiness) from sleepier days.

Again, I would like to examine any correlations between reported sleepiness and lapses in attention as well as with reaction times. But like the first two plots I have described, this one was just as much of a non-issue. It looked more like there was a correlation between reported sleepiness and lapses in attention. However, the reaction time information does not immediately seem to correlate.

That’s all for now. Expect more discussion of analyzing this data but don’t expect too many more days of actual polyphasic sleep. Due to starting a new job and having to feed and inject our aging diabetic and blind dog, Tiger, while in Calgary without changing his schedule, I will be transitioning back to monophasic sleep sometime in the next week. I realize that many people will say I’m copping out, or that my data is inconclusive because I didn’t complete the requisite month of adjustment but such is life. If I was seeing more drastic results in the data I think I would be keener to continue to pursue this method of sleep, but right now it doesn’t fit with my duties/lifestyle. I will however, after a week of recovery and a week of new baseline testing for the n-back task and some sleepiness measure, be reducing my sleep from 8 hours to 6 hours. This time the sleep will be in one large chunk, and will hopefully provide a nice contrasting data set to the one I have already built on polyphasic sleep.

So keep following this blog and, while it may be a bit more boring for a few weeks, it should pick up again when I start rambling like a sleep deprived person again!