Statistics and charts

So you’ve gathered all that sleep data – what do you do with it? Sleep as Android offers you easily readable reports for the quality of your sleep. Read about them in this chapter.

Where to find them

All statistics can be found on the stats tab inv_ic_action_stats. For charts tap the chart inv_ic_timelapse icon on the statistics screen.

Sleep deficit tracking 

Ideal daily sleep income is the second option in Settings.  This option represents a duration of sleep you’d ideally get every day. It is a very important tool to control your healthy sleep income. According to various sleep and life style studies a healthy sleep length for an adult is around 7-8 hours (assuming monophasic sleep, see polyphasic sleep). Ideal sleep income in Sleep as Android defaults to 8 hours.
Setting your ideal sleep income is the base for sleep deficit tracking presented in the Stats tab. There you will find your actual sleep deficit for a particular day, your total deficit for the observed time period (14 days, month, quarter…) and average deficit.
daily sleep deficit = daily sleep length – ideal daily sleep income
Selecting different time periods in the Stats tab allow you to see whether your sleep debt is improving or worsening. For example a smaller sleep deficit in last 14 days in comparison to your last quarter means you are doing well in the last 14 days.
Bedtime notification is an effective way to fight high sleep deficits or to maintain optimal sleep length. This feature may be customized in Settings > Bedtime notification. As the name suggests, time to bed notification notifies you when you should go to bed in order to not increase your sleep deficit.

Deep sleep %

After sleep length, deep sleep % is another very important measure which is related to sleep quality. Please refer to chapter How it works for more details on sleep graphs and the role of deep sleep %. In general a healthy deep sleep % should be in range from 30% to 70% assuming sleep tracking is applied correctly as described in Sleep tracking.
Deep sleep is very important in the overall role of sleep for our well being. Lack of deep sleep may indicate sleep issues such as sleep deprivation. You can compare your latest deep sleep development with your historic data to see whether your deep sleep % is improving or not. You can also see the effect of various life style events on your deep sleep using tagging described in the next section.

Tagging and rating

  • Adding tags
Each sleep graph allows adding life style tags. You can add tags in the morning dialog after you finish sleep tracking, in sleep graph detail using the pencil inv_ic_pencil icon or long press on a sleep graph and choose Edit record. You can also add tags ex-ante when starting sleep tracking. For example, you already know that you did #sport that day, you can add this tag right after starting the tracking. This is done in the sleep tracking screen using MENU > Edit.
For all available tags, see the lists below.
You may add any of these tags easily – they are available as icons in the rating dialog:
  • #food
  • #sport
  • #alcohol
  • #stress
  • #caffeine
  • #love
  • #med
  • #dream
  • #talk
  • #snore
  • #laugh
  • #sick
  • #work
  • #cpap
  • #baby
  • #note
  • #gooddream
  • #baddream
The tags below are being used automatically by Sleep according to data gathered in the specific sleep tracking record:
  •  #newmoon
  • #fullmoon
  • #rain
  • #stormy
  • #cloudy
  • #cold
  • #hot
  • #watch
  • #geo1
  • #geo2
  • #geo3
  • #geo0
  • #home

Analysing tagged sleep

Tagging allows you to study your sleep measures in relation to your tracked life style events. For example with tagging you can discover the effect of caffeine, alcohol, sport, food income etc… on your sleep.
This can be done in the Stats tab (in the image below). There is a collapsible Tags section which shows average sleep measures so that you can easily compare your overall average measures (sleep length, deep sleep, snoring…) with your tagged results. If you observe for example significant drop in deep sleep % after using caffeine you may consider to reduce caffeine consumption (especially before sleep) in order to improve your deep sleep %.
  • Coloring
We use green for surplus and red for deficit on sleep length.
For deep sleep % we calculate the average value:
    • if +-15% from the average the coloring is green
    • for any unusual values out of the average we use red.

Values above the average may mean some error of measurement – for example the phone was not close enough to the body. Values under average may indicate too much light sleep due to various factors such as stress, alcohol consumption, loud environment etc…

Another insight into tagging is provided in Charts, which is discussed in detail in the next section. There you can see effects of life style events in various bar charts. For example the effect on snoring length:
device-2013-11-08-124452
  • Custom tags
You can either use predefined tags which include the most common lifestyle events which have some relevance to sleep, or you can invent a set of your own tags. The only requirement is using the hash mark notation at the start, followed by any number of alphanumeric character (For example #headache or #travel).
  • Frequency tracking
It is also possible to track frequency of your events. For example #caffeeine_3x – meaning 3 coffees, #alcohol_2x – 2 shots, #sport_3x – 3 hours of sport. This allows to describe your lifestyle events more precisely. Support for automated analysis of the frequency information is not yet included in the app, but it is planned in coming versions.
You can add frequencies easily but pressing the tag button multiple times. To delete a tag on the other hand — long press the tag inv_ic_tag button
  • Rating
Rating is another quality criterion which adds subjective evaluation of our sleep. Similarly to deep sleep % we can show what sleep measures lead to your higher satisfaction with your sleep. Rating is added as 0-5 stars with 5 starts being the best rating. Sleep can be rated on the same dialog which is also used for adding tags.
  • Location tagging – geo tags
We use network-based location to implement some of the cool location based features such as wake up with sunrise, location tagging, and morning weather.
In the geotagging feature, we calculate the 4 most frequent places from your sleep tracking history, so the most-frequent place is considered “home”, the second one is “geo1”, then “geo2” and finally “geo3”.
The location we use is coarse and we store it even coarser, +/- 5 km. In effect, we identify your location as a square and then match those squares to set up the location tags.
You can completely turn off location storing in Settings > Miscellaneous.

Charts

Charts provide more insight into Your sleep data and put them into perspective.

You can filter your data by period (Week, 14 days, Month, Quarter, Half year, All), by days (Workdays, Weekend) and by tags. We recommend to use longer period rather than short in order to get statistically significant data into your charts.

Tab:Duration inv_ic_action_time

Duration chart

This line chart shows you the evolution of your sleep lengths. On the left, the chart starts with a date you set with the bottom dropdown menu and then it goes all to way to the right, which is today.
The green line shows you how your whole sleep length has been changing, while the dark blue line shows only how your deep sleep length has been changing.The light blue lines are averages (one for whole sleep length, one for deep sleep length).For example, in the chart below, you can see how my sleep was oscillating, but the average is still around 7 hours. You can observe a drop in deep sleep length on 3rd February.

How should it look? Ideally the chart should not have many extremes (regular, periodic chart is normal) and on the average, deep sleep should be ca half of whole sleep duration.

screen15_04_06

Tab: Deficit inv_ic_action_bedtime

Deficit chart

This chart is a useful chart for tracking Your sleep deficit. There are two lines.
The green line shows the deficit on a given day. If You sleep less then Your ideal sleep time, this line will be negative on that day. If You sleep more, it will be positive.The red line on the other hand shows the total deficit. Every day’s deficit gets added to the deficit pool, so You can see here, how big the actual sleep debt was. This value is what You have probably felt the most – whether You were refreshed or totally exhausted.

screen15_05_15
screen15_05_36
screen15_04_49

How should it look? You should strive to have zero sleep deficit, so ideally both lines, green and red, should be very close to 0 hours all the time.

Tab: Graphs inv_ic_action_graph

Sleep graph

This chart presents another view on Your normal sleep records. You can observe Your sleep patterns here, i.e. at what time do You go to sleep usually, at what time do You wake up and if You choose short enough period (e.g. a month), You can inquire into Your light sleep/deep sleep patterns.Green parts are light sleep.Dark blue parts are deep sleep.

screen15_05_57

screen15_06_11

screen15_05_47

Tab: Trends inv_ic_action_stats

On this tab, You can see the evolution of Your sleep characteristics all in one place.

Deep sleep % chart

This chart shows You how Your deep sleep has been changing recently – effectively how well did You sleep.

screen15_06_22

How should it look? The average should be kept around 50% (+- 15%). Ideally every day, You should have a 50% deep sleep on a normal sleep pattern (i.e. sleeping 7 hours/day).

Snoring chart

Here You can see the evolution of Your snoring time per night. If You want to see what has an effect on Your snoring, use the tags everyday and then look at the Tags charts tab.

screen15_06_27

How should it look? There is no single answer – e.g. if you wake up people around you, it may be time to start worrying about the percentage. However to snore a little is perfectly normal. So this is up to you and your surroundings.

Rating chart 

You can see the history of how did you rate Your sleep in the selected period.

How should it look? If you don’t rate your sleep at all, this will be zero. Otherwise, you will see a jagged line.

Fall asleep hour chart

You can compare the times at which you have been going to sleep recently in this chart.

screen15_06_35

How should it look? It depends, but generally abrupt changes in the time at which one falls asleep may have a negative impact on the body and energy levels.

Smart wake up chart

This chart shows how many minutes before your set alarm have you been waking up (while using the smart wakeup feature). This interesting measure visualized in charts is wake up before scheduled alarm time, which basically tracks what difference smart wake up does for you and whether your body is accommodating to smart wake up or not. This stems from an interesting theory that your body is able to adapt to the smart wake up feature over time and pretend that you are in deep sleep (not moving) before the alarm to increase sleep time.

Tab: Tags inv_ic_tag

In general, these charts connect the tags with sleep characteristics. You can e.g. track whether sport has some effect on your deep sleep percentage, whether you snore more after alcohol and gain other useful insight. You have to use tags of course in order to benefit from these charts.

Deep sleep chart based on tags

screen15_14_34

Duration chart based on tags

screen15_14_07

Snoring chart based on tags

screen15_14_25

Tab: Chronotype

Chronotype is a term used to characterise your sleep patterns. It is a spectrum ranging from 100% night owl to 100% morning lark. Read our introduction article.

Your currently detected chronotype will be always shown in the left menu. This one is detected from the last half year of sleep data.5cee18bc-6f65-46fe-9316-62bb038fec58Sleep irregularity chart shows how stable your fall asleep hour, wakeup hour and sleep duration is. Most regular sleeper would have a horizontal line from left to right. The irregularity shown is the variance of your mid sleep hour.
24b06229-dd90-4d19-b031-5689369ca67eThe chronotype trend chart shows the evolution of your owliness/larkiness. This chart always shows all-time data since shorter period do not have much meaning to the trend.

The chronotype value (mid sleep hour) is computed from weekend or “weekend-like” days to get the natural chronotype (see http://sleep.urbandroid.org/documentation/extended-features/chronotype/).
04989d9b-1deb-4ff2-a2a3-c6ea9b3e2c5dThe mid sleep hour chart shows the average middle hour of your sleep records. This is a good indicator of your general tendency towards being a night owl or morning lark.

You can also see here where you stand among the general population. The population data are computed from SleepCloud.
c69f6d8a-cfbc-4298-8316-e3db8a949b05

The social jet lag chart shows how your sleep pattern differs on workdays and on weekdays. We actually do not use the normal definition of workdays and weekdays, but compute “work-like days” and “weekend-like days” from your historical data.
e56ce801-3930-4cb3-a9f5-50208d65ff57

Tab: Advice inv_ic_lightbulb

If you don’t know how could you make your sleep better, go for some advice to this tab. Sleep as Android gives you evidence-based advice along with supportive charts. Of course such advice needs to be taken relaxed and used only as a slight guidance in which direction you may want adjust your sleep habits. But we don’t recommend to do any dramatic changes to your sleep. Our recommendation is gradual iterative adjustment with feedback from the sleep tracking. Also the performed analysis always needs some correction and interpretation from the user side.

bc84bd57-95e8-4027-b5d0-446f0ae7a6fc

Advice regarding your social jet lag

Fall asleep hour – regression model 

Here you get a variation on the Fall asleep hour chart with a twist. You can observe the influence that you fall asleep hour has on the quality of your sleep (deep sleep % and your subjective rating). Sophisticated methods are used in order to extract more useful data. You can see a regression model. This means that we have processed your data, filtered out any outliers (i.e. extreme values) and found a mathematic formula that best fits your data. On the basis of this formula, it is possible to statistically predict future behavior.

screen15_07_08

Duration – regression model

Similar case to the fall asleep regression model, you get a polynomial interleaving your filtered data. The data are presented in a way that you can see what influence has the duration of your sleep on its quality.

screen15_07_19

Weekday charts

There are two weekday charts – on one you can see on what days in the week you sleep more, i.e. your sleep lasts longer. On the other, you can see on what weekdays you have better quality of sleep (better deep sleep %).

screen15_07_27

Fall asleep charts 

There are three fall asleep charts in the Advice tab. You can see what effect your fall asleep hour has on sleep duration, deep sleep % and the amount of snoring.

SleepCloud

SleepCloud is a web-based service which allows you to backup and share your sleep data among several different devices. But more importantly, SleepCloud aggregates sleep data from many Sleep as Android users in order to find new information on sleep.
An important aspect of SleepCloud is we get information about the average sleep measures for thousands of Sleep as Android users. In other words you may compare you sleep with others finding out what is normal and what is extreme.
For example at the time of writing this documentation those where the current SleepCloud data:
Amount of sleep graphs in SleepCloud
500 thousand sleep graphs
Average sleep length
7:03
Average deep sleep
58%
Sleep cycles
4
Average snoring duration
22 minutes
You can find most recent data at http://sleep-cloud.appspot.com/.
We have also build in several analytic features into Sleep Cloud. You can watch trends, relations between two different sleep measures and see what are different sleep habits for people in different countries. Please see the screenshots below.
screenshot-from-2013-11-08-13-42-32
screenshot-from-2013-11-08-13-41-47
Share