The one change anybody logged

The before and after

Light endings per sitting: 0.36 before, 0.06 after. Median sitting length up thirteen minutes. For another evidence base on people, work, or environments, see CSIRO.

On the face of it, the clearest result on this site.

Why it is not

Six reasons, and any one of them would be enough.

It is one household. Nine dwellings is already too few; one is not a number.

It is eleven sittings against sixteen. Four light endings and one light ending. A difference of three events.

Nothing was controlled. Wednesday to Friday was a different part of the week, with different work in it, and the household's own notes mention a quieter second half.

The weather is unknown. The sheet did not record it. Three days with more cloud than the first two would produce exactly this result with no furniture having moved at all.

The household knew they were logging. Somebody who has just rearranged a room in response to a log they are keeping has an interest in the rearrangement working, and the field they would express it in is the one they fill in themselves.

It was not planned. This was not an experiment with a before and an after. It was a person getting annoyed on a Tuesday and moving a table, which happened to fall inside a week somebody was counting.

Three fewer events in three days, in one dwelling, with no control, no blinding, no weather record and a motivated observer.

Why it is on the site at all

Because it is the only intervention in the entire record, and a site that reported ninety-four endings while omitting the single occasion anybody did anything about one would be reporting selectively.

And because the direction is worth noting even if the size is not. The household moved a place away from the thing that was ending its sittings, and the sittings stopped ending that way. That is the mechanism working as the rest of the record would predict.

It is the mechanism behaving as the rest of the record would predict, not a result.

What a real version would look like

Two dwellings, matched as closely as nine candidates allow. One moves a place, one does not, both log for four weeks with a weather record alongside.

That is a small, cheap, entirely feasible study and it is not this one. It would answer a question this record raises and cannot touch: whether a household acting on the largest cause in the data can shift it.

What the household said afterwards

That the new position was worse to sit in and they moved it back the following week.

That single sentence is the most useful thing in this entire page, because it describes a trade the record has no field for. The rearrangement reduced light endings and the household disliked the result for reasons the log could not record, and they reversed it once nobody was counting.

A measure that captures one side of a trade will always show improvement when somebody takes the other side of it. This record captured endings and missed everything the household was weighing against them.

The lesson for the rest of the site

Every finding here is a count of one kind of event. Wherever a count of one kind of event is used to judge whether an arrangement is better, it will be wrong in the direction of whatever it counts.

That is why no page on this site says any dwelling, place or arrangement is better than any other — and why this page, which has the only before-and-after in the record, concludes with the household putting the table back.

The move itself

The place was a table in a spare room, against a wall with a window to one side. It went to the opposite wall, facing into the room.

Four light endings in two days became one in three. The median sitting rose from forty-four minutes to fifty-seven. The household's other endings — two wanted, one noise, one surface — were unaffected, which is at least consistent: moving a table away from a window should change light endings and nothing else, and it did.

That specificity is the only thing about this result that is hard to explain away. Cloud cover would have done the same, and a quieter second half of the week would not have touched the other three rows either.

It remains one dwelling, three days and three events.