Creating a Self Administered Sleep Scale

These last few weeks have been a symphony of horror when it comes to sleep. I thought I had this nailed down, but in getting back from Spain from the States my sleep quality and quantity has been all over the place. This negatively affects almost every other habit I have, and it's a problem I've had for a very long time.

There are many variables I'd like to test out. The biggest problem I've encountered is finding a good metric that I can test variables against. A quick search resulted in this great overview. It outlines several scales, including:

  • The Medical Outcomes Study (MOS) Sleep Scale
  • The Insomnia Severity (IS) Index
  • Pittsburgh Sleep Diary (PghSD)
  • Pittsburgh Sleep Quality Index (PSQI) Instrument

What I want is a simple self-administered scale (just like the SRHI for habit formation) that I can use as a base to deliberately test out behaviors to see what will affect my sleep. There is so much advice out there, but almost none of this is systematically tested. I've also had several friends who have gone in for sleep studies only to get vague advice in return, usually related to sleep apnea. 

I found that the Pittsburgh Sleep Diary (PghSD) and the Pittsburgh Sleep Quality Index (PSQI) Instrument both include too much detail and contain behavioral questions that I want to test out personally (how much alcohol did you consume, etc). I found the Insomnia Severity (IS) Index more feeling-based and difficult to assess with any accuracy over stretches of time. I found the Medical Outcomes Study Sleep (MOS) Scale good if applied to a week rather than the 4 weeks time periods questions were based on. The best thing I found was one section of the PSQUI on Habitual Sleep Efficiency (HSE). HSE is a simple weekly percentage:

# hours slept/# hours in bed X 100

I find this very elegant, but it does not assess quality of sleep. 

Several questions spread out across multiple scales addressed sleep quality, so I decided to just collect them all irrespective of scale. Some had to do with daytime dysfunction (naps, sleepiness, alertness, enthusiasm), sleep disturbances (how many times did you wake up), and sometimes data collected from sleeping partners. 

Many of these questions dealt with quality of sleep during a month. Personally, I believe that remembering a month of sleep is not only inefficient, but highly inaccurate - I have problems remembering what happen a few days ago. That being said, I believe that assessing variables based on on one week would speed things along. Under the Likert-type scale, answers are based on sliding strengths of opinion. The SRHI is a great example of this with it's 7 point answer system, though many of the sleep scales have a 5 option system.

Here are some questions I collected. I include this to show examples of not only the types of scales but how to potentially create self administered scales in the future, as I find them highly useful in productivity and personal development.

(All of the time/most of the time/a good bit of the time/some of the time/a little of the time/none of the time = 1, 2, 3, 4, 5, 6)
How often during the last 4 weeks ...

...did you feel that your sleep was not quiet (moving restlessly, feeling tense, speaking, etc, while sleeping)? 
...did you feel that you got enough sleep to feel rested upon waking in the morning
...did you feel drowsy or sleepy during the day
...did you awaken during your sleep time and have trouble falling asleep again
...did you have trouble staying awake during the day
...did you take naps during the day
....did you get the amount of sleep you needed

  • From the IS Index:
    (0 = none, mild = 1, 2=moderate, 3 = severe, 4 = very severe)

-difficulty staying asleep
-how satisfied/dissatisfied are you with your current sleep pattern?
-how noticeable to others
-how worried/distressed are you about your current sleep problem
-to what extend do you think your sleep problems interfere with your daily functioning - fatigue, mood, ability to function at work daily chores, concentration, memory, mood

  • From the PghSD:
    (specific numbers/written descriptions)
    -how many naps?
    -how many times did you wake up during the night?
    -different types of waking up?
    -sleep quality - very bad, very good - whole scale between
    -mood on final awakening
    -alertness on final awakening
  • rom the PSQI:
    How many times do you...
    -wake up in the middle of the night or early morning
    -have to get up to use the bathroom
    -overall sleep quality
    -trouble staying awake while doing things during the day
    -enthusiasm to get things done 

I went through these and narrowed them down and modified them. There were several problems with this. First, I had to include multiple elements to the scale. The SRHI doesn't just include length of time with respect to habit formation. It includes length of time, automaticity, identity, etc.

Secondly,  the all had to be scored similarly. Some were on a 5 points scale, others were specific numbers (how many times did you get up during the night?). If there were differences, different types of questions would have greater or lesser weight. This seems really obvious, but I didn't think of it at first, and I think it's just something to remember in the future when creating metrics.

Thirdly, I had to figure out what time frame the questionnaire would be based on. Most of the exams I pilfered questions from were for a month. I don't think this is very accurate - I certainly don't remember my sleep patterns after a few days. So I'm going with a daily assessment with a weekly average in order to test out variables for a week. 


  1. [Sleep quality at night] How severe was the restlessness of your sleep?
    0=Very Severe, 1=Severe, 2=Moderate, 3=Mild, 4=None
  2. [Morning] How well rested do you feel in the morning?
    0=Exhausted, 1=Not Rested, 2=Moderately Rested, 3=Rested, 4=Energized
  3. [Rested, next day] How often do you feel drowsy or sleepy during the day?
    0=Incredibly drowsy, 1=Quite Drowsy, 2=Somewhat Drowsy, 3=A little drowsy, 4=Not at all
  4. [Sleep quality] How often do you awaken during your sleep and have trouble falling asleep again?
    0=Every Day, 1=Several Days, 2=A Few Times, 3=Once This Week, 4=Never
  5. [aps] How often do you take naps during the day this week?
    0=Every Day, 1=Several Days, 2=A few days, 3=Once a Week, 4=Never
  6. [General] How often this week did you get the amount of sleep you need?
    0=Never, 1=Once, 2=A Few Times, 3=Almost Every Day, 4=Every Day
  7. [Functioning] To what extend do you think your sleep problems currently interfere with your daily functioning?
    0=Every Single Day, 1=Almost Every Day, 2=A Few Times a Week, 3=Once, 4=Never

Right now, I see my sleep scale to be comprised of both the Habitual Sleep Efficiency (HSE) and the above modified sleep quality index. My intention is to take both every day to be averaged for the week. I'll be doing a controlled test this week, just to see how it works and the ins and outs of taking it. I'll check to see if some elements need to be modified.

Yesterday just happened to be a day that I had a particularly bad night's sleep. Since I already have a habit of sleep recording, it was pretty easy to figure out my HSE.

HSE = Total Hours Slept  (6) / Total Hours in Bed (11) X 100 = 54.5%

Sounds about right.

My, for lack of a better label, "Sukumaran Sleep Quality Index" was a little more difficult. Here were my results:
1. 0
2. 0
3. 0
4. I got up once
5. I took one nap
6. 0
7. 0

Numbers 1-3 were simple. Questions 6 and 7 could be modified to read "Did you get the amount of sleep you needed today?" and "To what extent did your sleep interfere with your functioning today?" respectively. Questions 4 and 5 were weird because though getting up once might be considered a medium response on the Likert scale, repeatedly having to get up over a week might result in a worse score. For question number 5, for a daily question my answer would most likely be binary - either I took a nap or I didn't. Added up for the week would result in a number that means something. So it's really difficult to take the scale accurately for a day.

It might just be best to keep all of this information, plus side notes - like the number of times I feel drowsy during the day (yesterday's total count was 32, tracked with my handy hand tally counter) - in a specified document, or appended to my daily log.