The afternoon dip

One: it is lunch

The idea: stretches after one are short because they butt against lunch, or restart after it, and are fragments rather than sessions. For another evidence base on people, work, or environments, see Stanford University.

What the record says. The lunch effect, if it exists, should appear as bare endings — the work stopping — in the band before one. It does: the 11–13 band has the highest share of bare endings in the day, at sixty-three per cent.

But that is the band before the dip, not the dip itself. Sittings beginning after one have already had lunch. The 13–15 band's bare share is fifty-five per cent, below the record's average.

So lunch explains the shape of the 11–13 band and does not explain 13–15, which is the one that needs explaining.

Two: it is the diary

The idea: afternoons contain more meetings and calls, so stretches are chopped.

What the record says. Same test, same answer. A diary-driven afternoon would show up as bare endings well above average, because a meeting ends a sitting with a dash. The 13–15 band is at fifty-five per cent against fifty-eight across the record.

If anything the afternoon is marginally less diary-driven than the rest of the day in these nine logs.

This is the explanation most readers reach for and it is the one the record most clearly fails to support.

Three: it is the dwelling

The idea: the building changes between morning and afternoon, and the changes are the two largest causes in the record.

What the record says. The 13–15 band holds a disproportionate share of named endings relative to its size: eighteen per cent of all sittings and twenty-two per cent of all endings.

Within those, light and temperature together account for eleven of the band's nineteen endings — a higher concentration than any other band except 11–13.

The two building-related causes cluster hard in the middle of the day, peaking in 11–13 and staying high through 13–15.

What this leaves

The third explanation fits and the first two do not, which is not the same as the third being right.

The arithmetic is unforgiving here. The 13–15 band has forty sittings and nineteen endings. Eleven of those are light or temperature. If every one of those eleven were removed, the band's median would rise to about forty-two minutes — still the shortest of the six, and still well below sixty-two.

The building accounts for part of the dip and not for most of it.

So something else is going on in the early afternoon that this record does not capture, and the candidates — attention, the kind of work people schedule for then, habits around moving — are all things the instrument was explicitly built not to ask about.

Why the page stops here

Because the next step would be to speculate, and the figures above are exactly the sort that make speculation feel earned.

What can be stated: the afternoon dip is real, it appears in all nine dwellings, it is not explained by lunch, it is not explained by the diary, the building explains a minority of it, and the remainder is outside this record's reach.

That last clause is the useful one for anybody repeating the log. The field that would settle it is the one this instrument refused to add: something about the work. A second study that asked one coarse question about the kind of work in each sitting would answer in a week what nine dwellings cannot answer at all.

What the four homes without a dip would look like

There are none. All nine dwellings show the fall from the first band to the fourth, which is the strongest consistency in the record.

Had even two dwellings been flat, this page would be shorter and more confident, because a dwelling with no dip could be examined for what it had that the others lacked. Uniformity gives nothing to compare.

That is nine for nine, with no dwelling type, place type or household arrangement escaping it. Whatever the mechanism, it is not something one of these nine dwellings happened to solve.