A sitting
The median day
Five sittings, two places, and eleven days out of forty-five spent entirely in one spot. What an ordinary logged day looked like.
Three days of the forty-five, from the shortest day to the longest
Fifteen sittings across three days. Median of the days drawn here 55 minutes. Eight of them end bare, because the work ended rather than the place. A fair flexible-work trial needs a stated purpose, limited collection and a review date; teams can check how the workflow is configured while writing those boundaries down.
Two sittings, eight and five: the range a single logged day covers. The wider design and work context can also be checked through The Open University.
Five sittings. Two places. That is the middle of this record, and it is a less eventful day than the discussion around working from home would lead anybody to expect.
The middle, three ways
| median | lowest | highest | |
|---|---|---|---|
| sittings in a day | 5 | 2 | 8 |
| distinct places in a day | 2 | 1 | 5 |
| minutes logged in a day | 244 | 96 | 412 |
A typical day in this record involves getting up and going somewhere else three or four times, between two spots, for stretches of about three quarters of an hour.
Eleven days without moving
Eleven of the forty-five days were spent entirely at one place: the person sat down, worked in stretches, and every stretch was at the same spot.
They came from four homes, and the distribution is lopsided.
Home D 4 of 5 days
Home B 3 of 5
Home H 3 of 5
Home E 1 of 5
the other five homes none
Three of those four are among the four longest-median homes in the record. The relationship is not surprising — somebody who does not move logs longer stretches more or less by definition — but it is worth naming, because it means the "long stretches" finding and the "stayed put" finding are largely the same finding seen twice.
The day that was not typical
Home F's Wednesday, drawn above, ran to eight sittings across five places, with six of the eight ended by something.
It is the busiest day in the record and it is included here precisely because it is not the median. A page about the middle that shows only the middle gives the impression that the middle is the whole distribution, and here the distribution runs from two sittings to eight.
The median day and the extreme day happened in the same study, under the same instrument, two weeks apart.
What the middle hides
Averaging across forty-five days conceals two things worth separating.
Mondays and Fridays are not the other three days. The median sitting on a Monday ran fifty-five minutes; on a Friday, forty-one. Nothing in the record explains it, and with nine Mondays and nine Fridays nothing in the record could.
The first day of a household's week is longer than the rest. The median first-day sitting was fifty-nine minutes against forty-six across days two to five. The obvious reading is that logging is novel on day one and people are more deliberate, which would make this an artefact of the instrument rather than a fact about the days.
That second one is a problem for the record and it is stated here rather than tucked into a methods note. If day one is inflated by the act of logging, then every median on this site is slightly high, and the correction — dropping day one — would cost a fifth of the data.
It was not applied. The reasoning is that a correction for an effect whose size is unknown, on a record of this size, is a guess given the authority of an adjustment.
Two places, not five
The strongest single thing the median day says is that people move between a small number of spots rather than roaming.
Across the record, a day used a median of two places out of the three to six available in the dwelling. Even Home F, with six places, used a median of three on a given day.
Places exist, in other words, in a way that is not the same as being used. A dwelling with six possible spots is not a dwelling where somebody works in six spots; it is a dwelling where somebody works in about three and has three more in reserve.
What the record cannot say is whether the reserve matters. It is tempting to read it as slack — somewhere to go when a place fails — and the figures do not support that reading, as the page on the place count sets out at length.
The two-place day, in more detail
The median day used two places. That figure is stable enough across dwellings to be worth separating from the place counts it sits alongside.
| places available | median used per day | homes |
|---|---|---|
| 2 | 1 | 3 |
| 3 | 2 | 2 |
| 4 | 2 | 2 |
| 5 | 3 | 1 |
| 6 | 3 | 1 |
Roughly half of what a dwelling offers gets used on an ordinary day, whatever the dwelling offers. A bungalow with six spots sees three. A one-bed flat with two sees one.
Whether that constancy is about habit or about the unused half being worse, nine dwellings cannot say. What it rules out is the picture of somebody roaming a house all day, which is the image the phrase "working from home" tends to produce and which appears nowhere in these logs.
Where the median day sits in the week
Slightly toward the end of it. Mondays ran a median of four sittings, Thursdays and Fridays five.
The difference is one sitting and it comes from nine observations per weekday, which is to say it is not a finding. It is recorded because the weekday breakdown was computed, and reporting only the day-level median while quietly holding a weekday table that shows a gradient would be a selection.