Subjective Sleep Score vs Oura Ring Metrics
How does subjective sleep assessment compare to Oura metrics? My recent short tracking experiment can help me to answer this question.
For the past 3 months, I have been tracking two additional metrics in my daily log. Upon waking up, I would ask myself “how well did I sleep”, and log my subjective sleep score using two scales in rTracker.
The first scale (I call it qualitative) used a simple 5-point scale. The second scale was a sliding scale with values from -5 to 5:
The main question I am trying to answer is how close does subjective sleep assessment come to Oura’s objective metrics, and which ones. Over three months of logs was enough to produce some reliable correlations to answer these questions. Naturally, I had to normalize and detrend the data.
First, let’s look at Oura metrics related to various sleep stages:
Not suprisingly, duration of sleep was positively and (relatively) strongly correlated with subjective sleep score. Interesting, subjective sleep score was also positively correlated with light and REM sleep, but not with the deep sleep.
I can also look at the various biometrics tracked by Oura during the sleep, like heart rate, breathing rate and body temperature:
As you can see, none of the biometrics is correlated with subjective sleep score. I actually did not expect any patterns there.
Finally, Oura offers some composite “summary” scores that I thought would be also interesting to look at:
Both Oura’s Sleep Score and Readiness Score are positively correlated with my subjective sleep numbers.
Based on these results, I would conclude that subjective sleep scores can be a good way to validate Oura’s metrics, but not sensitive enough to differentiate betwen various dimensions of sleep. Also, if one still chooses to track sleep subjectively, a simple 5-point scale should suffice.