Knowing our alertness level

The measure of reaction time is known scientifically as the best clinical tool to evaluate alertness. This must be done in a specific manner as done with the PVT (Psychomotor Vigilance Task). Sleep-2-Peak’s reaction time test has been validated scientifically as equivalent to the PVT or even better for some aspects. It is also proven that personal impression to be fully alert is often false and that reaction time tests allow to demonstrate fatigue not perceived by ourselves.

Typing speed («texting speed», «Texting Intercharacter Time (TICT)» or «keystroke time») has also been demonstrated to vary with alertness levels. For example, the group of Althoff et al. from Stanford University and Microsoft have demonstrated that keystroke time on a PC keyboard after one night with less than 6 hours of time in bed took 4 days to recover and that after 2 nights with less than 6 hours of time in bed it required 7 days to get back to normal.

Link between reaction time, texting speed, performance and wellbeing

During my extensive study project looking at correlations between sleep and performance, it has been shown that personal variations in physical performance were highly correlated to variations in reaction time. Cognitive performances and wellbeing feeling are also strongly associated to the alertness level. This is well supported by numerous studies looking at the effects of sleep deprivation.

How to use the app

Sleep-2-Peak allows to evaluate alertness levels associated to performance with two measures: reaction time and texting speed. For reaction times, the basic test lasts 2 minutes and includes 12 trials. The test duration can be modified if wanted. Texting speed is measured with a specific algorithm using a baseline texting speed measure for that user. A brief calibration of Text-speed is required when starting this function for the first time.

The results for reaction times and for texting speed are then linked on the first graph to the bedtime and out of bed time of the night preceding the reaction time tests or the texting speed measure. By accumulating your results, this allows to visualize which bedtime and out of bed time are associated to your best performances the next day. An average curved line is drawn between results an hour or less apart.

s2P, graph 1

A second graph allows to see daytime performance variations (circadian variation). Results from the last 24h are illustrated with bullets. An average curved line of all accumulated results (or those defined by the filters) is drawn between results separated by an hour or less.

s2P, graph 2

To discover your optimal sleep schedule and see if you reach your highest possible performance levels during the day, here are the basic recommendations:

For reaction time tests, prioritize the following moments:

  1. When you feel tired in the evening but not yet drowsy. From our studies, this moment corresponds to approximately 60% of your maximal alertness. This moment is an important reference point to determine the amplitude of your daytime alertness.
  2. When you feel at your best during the day. For most individuals, this will be a few hours after awakening in the morning. See if you can reach 30ms faster than your reaction times when tired in the evening.
  3. Try to do other reaction time tests at different times of the day and in different contexts to better discover your alertness level variations.

For texting speed, this is measured automatically when a sufficient number of characters are taped in a one hour period, representing in general a few lines of text respecting the algorithm. The average will be referred to the precise hour at the center of that period. For example, texting speed for 11h00am represents the average for texting done between 10h30am and 11h30am. For iPhones, it is important to use the sleep-2-Peak keyboard for the texting speed to be measured. The more you text at different moments of the day, the more your alertness curve based on your texting speed will be defined rapidly.

A possibility to see rapidly your alertness curve based on your reaction times and texting speed is to devote a 24 to 48h period where you will write a few lines of text and do a reaction time test at the switch of every hour from the time you get out of bed all the way to your bedtime. If you start your tests in the hours preceding your usual natural awakening and continue your tests past the appearance of drowsiness in the evening, you will be able to see even more obviously, and sometimes in a surprising way, the deterioration of your capacities. This also allows to better define your optimal alertness period.

Results interpretation

Best bedtime and out of bed time

By accumulating progressively your results, you will be able to see in general what bedtime and out of bed time are associated to your best performances the next day. This can be visualized with reaction times and/or with texting speed. If the results are different between these two measuring tools, it is possible that this is due to an insufficient quantity of results and that this will harmonize itself with additional data collection.

It comes back to you to explore different options of time in bed to find your best schedule. For example, if your best results up to now are for your nights with a bedtime at 11h00pm and an out of bed time at 7h00am, compared to your nights with a bedtime at 00h00am and an out of bed time at 8h30am, see if a bedtime at 10h30pm and an out of bed time at 7h15am could bring even superior results the next day.

Bring also attention to your other associated measures including if your awakenings were natural or not (alarm clock or forced to get up) and your active substances like caffeine. You can see the difference by using the filters to make your graphs. For example, compare your performances when you got up naturally vs when forced to get up, or with or without caffeine. If you take caffeine, does it allow you to have a nice and regular wake-sleep curve, does it bring an important deterioration at the end of caffeine effect (caffeine crash) or does it interfere with fatigue and sleep appearance in the evening?

Best times of the day for peak performance

The graph with performances based on time of the day allows to become aware of many aspects.

At what time do I reach a high level of performance? After you get up, a period of at least two hours is often necessary to reach your high levels of performance, this initial period usually being called sleep inertia. This can vary with different factors including if your awakening was natural or not, if you have an accumulated sleep debt and your own baseline circadian rhythm.

Do you have moments of increased or decreased performance during the day? See if in general you have moments during which you are particularly alert. These periods could represent good moments during which you could accomplish tasks requiring a high level of physical and/or mental performance. Inversely, do you have periods of decreased performance during which easier tasks would be preferable or even a nap could be appropriate?

Do today’s actions up to now allow you to reach better performance than usual, or to the contrary harm your performances? For example, are you jet lagged or just after a time change, have you started back a work or school schedule after vacations or an off period, are you sick, have you changed something in your diet, have you started medication or other substances? See if your performances are in general superior or inferior to your usual results.

Using results to improve your performances and wellbeing

Sleep-2-Peak can thus allow you to realize what bedtime period and what daytime moments allow you to be at your best, for your performances and wellbeing. For many, this will be a new awareness level since subjective perception of alertness is limited. A user’s data and graphs can allow the user to review his work/school/activity schedule while having objective data if needed to show others (employer, teacher, coach, manager, parent) what contributes to this user’s peak performance.

Sleep-2-Peak to bring societal changes for everyone’s wellbeing

For users accepting to share their results when the database will be established (crypted and anonymous results), these data will be able to eventually serve to better define and implement schedules favoring everyone’s wellbeing, for example by guiding leaders and decision makers to establish new work and/or school schedules.