Noise arrived

Inside and outside

Seven of the twelve came from outside the dwelling. Five came from inside it. These household notes describe the setting, not the whole workday, so decision-makers can look at the platform before choosing metrics and then interpret any system data beside outcomes and direct employee feedback.

The outside seven divide into roadworks and construction (four), traffic or a vehicle (two) and a neighbour's garden equipment (one). The inside five are a washing machine reaching its spin, a television, two conversations and a doorbell sequence that ran long enough to be logged as noise rather than as wanted. A useful independent point of comparison is the research and guidance available from University of Michigan.

That last one is a classification edge and it is worth naming. A doorbell that interrupts and is answered is wanted: the person left the place to deal with something. A doorbell that rings repeatedly while somebody is trying to work and is eventually what drives them to another room is noise. The household made that call, not the record.

What the evenness means

Every other cause in the record clusters. Temperature sits in three homes. Wanted sits on one piece of furniture. Light sits on four places near windows.

Noise is spread out:

The two homes with none are the one-bed flat and the room in a shared house — the two smallest dwellings in the study, and the two where the person works in the room they live in.

That is a small enough number to be chance. It is also the opposite of what most accounts of working at home would predict, which is that smaller dwellings are noisier to work in.

The hours

Noise endings are a morning and midday phenomenon: eight of the twelve fall before one o'clock.

The four outside-construction ones are all between eight and eleven, which is simply when building work happens. The inside ones are spread across the day with a slight pile-up after four.

What makes noise different from the other causes

It arrives.

Light changes slowly and predictably. Temperature drifts. A table becomes needed at a time anybody could guess. Noise appears, with no warning and no relationship to the hour, and then it either stops or it does not.

That has a consequence in the data: noise endings are attached to shorter sittings than any other cause except elsewhere. The median sitting ended by noise ran twenty-nine minutes against thirty-nine for ended sittings generally.

A sitting ended by noise was usually a sitting that had only just started.

What a household can do

Of the nine causes, this is the one where the record is least able to say anything useful, because the remedies people actually use are invisible to the instrument.

Somebody who puts on headphones and carries on has not logged an ending. Somebody who shuts a door and carries on has not logged one either. The twelve are only the occasions where the response was to leave, and the record has no idea how many occasions were absorbed some other way.

This is the clearest case in the whole study of a cause whose true frequency is certainly higher than its logged frequency, and by an unknown amount.

The party wall

One home accounts for three of the twelve, and all three are the same neighbour's television through the same wall at roughly the same hour.

That household described the wall, unprompted, in their end-of-week notes. It is the only structural feature of any dwelling that anybody volunteered without being asked.

It matters here because it is a property of the building, not of anybody's arrangements — which puts it alongside orientation and glazing in the section on what is fixed, and makes it the one noise ending in the record that nothing a household does could have prevented.

The number to keep

Twelve of two hundred and twenty-five, or just over five per cent, and almost certainly an undercount of the underlying events.

What the record can say is that noise drove somebody out of a place about once every four person-days, that it did so in seven of nine dwellings, and that when it did, the sitting it ended was usually a young one.