The hour
What the hour does to the endings
Not just how long a sitting runs, but what stops it. Each of the nine causes has an hour, and three of them barely overlap.
Named endings by the band the sitting began in
Six bands of the working day, built on two hundred and twenty-five sittings. Longest in 11–13 at 23 minutes, shortest in 17–19 at 7. A difference of 16 minutes across the same places and the same people. When the same room performs differently across the day, managers can compare the reporting options here and compare time records with the physical conditions logged here instead of guessing from presence alone.
Ninety-four endings, concentrated in the middle of the day. The wider design and work context can also be checked through University of Edinburgh.
The hour changes how long a sitting lasts. It also changes what ends it, and the second effect is sharper than the first.
Endings by band
07–09 11 endings of 38 sittings 29%
09–11 19 of 47 40%
11–13 23 of 41 56%
13–15 19 of 40 48%
15–17 15 of 36 42%
17–19 7 of 23 30%
A sitting that starts between eleven and one is nearly twice as likely to be ended by something as one that starts before nine.
Each cause has an hour
| cause | peak band | share in peak |
|---|---|---|
| light | 11–13 | 8 of 24 |
| temperature | 07–09 and 09–11 | 9 of 18 |
| wanted | 11–13 | 9 of 14 |
| noise | 09–11 | 5 of 12 |
| surface | 07–09 | 3 of 9 |
| elsewhere | spread | — |
| seat | 13–15 | 4 of 6 |
| power | spread | — |
Temperature is a morning phenomenon. Light and wanted are a midday one. The seat, such as it is, is an afternoon one.
The three that barely overlap
Temperature, light and wanted are the first, second and third largest causes after the bare end, and they occupy different parts of the day almost cleanly.
Nine of eighteen temperature endings are before eleven. Fourteen of twenty-four light endings are between nine and one. Nine of fourteen wanted endings are between eleven and two.
Laid end to end they cover the working day in sequence: cold first, then sun, then other people.
A dwelling that is cold at nine, sunlit at twelve and wanted at one is not three problems. It is one dwelling, running through its day.
What this does to the idea of a good place
It makes "a place that works" a statement that has to specify a window of hours, and the windows for the three main causes are different.
A back room that solves the wanted problem introduces the temperature problem. A spot away from the window solves the light problem and is usually further from the heating. The record contains at least four places where moving to avoid one cause moved somebody into another, and the endings log shows it: a temperature ending followed within the hour by a sitting that ends with light, and the reverse.
Six such sequences appear in the record. Six in forty-five days is not a pattern, but it is the mechanism by which a dwelling with several places can still produce a lot of endings — which is what the page on place counts could not explain.
The afternoon that has nothing in it
The 15–17 band has fifteen endings in thirty-six sittings, and no cause is concentrated there except the four seat endings.
By three in the afternoon the sun has mostly finished with these dwellings, the household has not yet come home, and the heating has caught up. It is the quietest part of the day for the building, and it is not the longest band.
That gap — quiet building, middling stretch lengths — is the clearest single piece of evidence that the hour effect is not only about the dwelling. If the building were the whole story, 15–17 would be among the longest bands. It is fourth of six.
The caution
Every figure on this page divides ninety-four endings into six bands and then into nine causes. Most of the resulting cells hold fewer than five observations.
The three large patterns — temperature early, light and wanted at midday — rest on nine, fourteen and nine entries respectively and are the only ones worth any weight. The rest of the table is printed because leaving it out would be a selection, and read as a sketch.
The bare endings have an hour too
Named endings cluster at midday. Bare endings cluster at the two points where a working day naturally breaks.
07–09 27 bare of 38 71%
09–11 28 of 47 60%
11–13 18 of 41 44%
13–15 21 of 40 53%
15–17 21 of 36 58%
17–19 16 of 23 70%
The curve is the inverse of the named-endings curve, which is arithmetically inevitable — every sitting ends one way or the other — and the shape is still worth seeing.
The 11–13 band is the only part of the day where the place ends more sittings than the work does. For three hours in the middle of the day, these nine dwellings are the leading cause of somebody getting up, and for the other nine hours they are not.
If this record has a difficult hour, it is not the afternoon. It is the three hours either side of noon, where the sun arrives and the table is wanted at the same time.
That is a narrower and more useful statement than the one the length chart alone would suggest, and it comes from reading the endings rather than the durations.