Fourteen-Day Sleep Diary
Record sleep timing, awakenings, naps, caffeine, exercise, and morning alertness for two weeks, and see the pattern underneath the individual nights.
The method is published. The tool is not built yet.
A calculator that will not show its arithmetic is not evidence-first. The assumptions behind an estimate are the part you most need in order to judge whether the estimate is worth anything, so on Restadian they are published first and separately.
That is what this page is. It sets out what the tool will measure, what it will ask you for, what it will assume, and where the method is still being specified. When the tool ships it will do what is described here — and if the method changes while it is being built, this page changes with it.
There is nothing to run on this page. Nothing written below depends on the tool existing yet.
The method
What it will record
A few entries a day, most of them one tap. When you went to bed, roughly when you think you fell asleep, when you woke, how many times you remember waking, whether you napped, when you last had caffeine, when you exercised, and how alert you felt in the late morning.
Fourteen days is the target because that is roughly the span over which a pattern becomes visible above ordinary night-to-night variation. A shorter record is still usable; it is just noisier.
What it will give back
A view of your fortnight rather than a verdict on any night: how much your bedtime and wake time moved, where weekday and weekend timing diverged, and what tended to precede your better and worse mornings.
The output is descriptive by design. It shows you your own schedule. It does not grade it, and there is no score to improve.
Where your data will live
In your browser, on your device. The diary will be local-first: entries are stored locally, nothing is transmitted to Restadian, and there is no account to create.
The practical consequences are worth knowing in advance. Clearing your browser storage will clear your diary. Your entries will not follow you to another device unless you export them. Export and import will be provided so you can keep your own copy and take it to an appointment.
The assumptions it will rest on
- Self-report is imprecise and still useful. Nobody can report their own sleep onset accurately. The diary is built around patterns across days, where that imprecision largely averages out, rather than around any single night’s numbers.
- Consistency of recording matters more than accuracy of recording. Estimating your sleep onset the same rough way each night produces a usable trend even if every individual estimate is off.
- The reader is the analyst. The diary surfaces the pattern; it does not interpret it for you.
Where the method is still being specified
The exact set of prompts is not final. There is a real trade-off between asking enough to be useful and asking so much that people stop filling it in after four days, and we would rather resolve that with a working prototype than declare it now.
What it will not do
- It will not assess, score, diagnose, or label you or your sleep.
- It will not tell you how long you actually slept, because it cannot know.
- It will not detect breathing, movement, or sleep stages. If something happens in your sleep that you are unaware of, a diary will not find it — that is what another person’s observation, or a sleep study, is for.
- It will not require an account or an email address.
Why it is still worth keeping
Because the questions a clinician asks are almost exactly the questions a diary answers, and almost nobody can answer them accurately from memory. Two weeks of imperfect records beats a confident summary of a fortnight you cannot recall.