The Survey / Self-Report Layer: What People Say, What Conventions Dominate, and What Neither Can Show

This document surveys, alongside the experimental dose–response evidence, a second, quite different body of material that circulates for the same four settings: what people say they prefer, what they are observed to do inside the schedules they already inhabit, and what institutions have standardised for reasons of timetabling, billing, and labour negotiation rather than performance. Traced to source, that material looks like this. The Pomodoro Technique's 25-minute work block and 5-minute short break (Francesco Cirillo, late 1980s) rest on one student's personal trial-and-error with a tomato-shaped kitchen timer, not on any experiment or survey. The DeskTime 52/17 pattern (52 minutes of tracked work, 17 minutes away, among the top 10% most "productive" users in 2014) is observational app-usage data selected on the outcome, not a preference survey and not a trial — and DeskTime's own re-runs moved the pattern to 80/17 in February 2020 and 112/26 in 2021, which is incompatible with treating the earlier value as a stable, context-free optimum. The recurring ~50-minute unit in classrooms (the Carnegie unit's 50–55-minute period), in psychotherapy (the 50-minute hour), and in lesson and appointment scheduling is a structural convention; no verifiable survey was found that asks knowledge workers for an ideal focus-block or break length and returns a defensible number. For inter-set rest, coaching tradition long prescribed the opposite direction from the experimental plateau — short rests for hypertrophy (30–90 s in the National Strength and Conditioning Association's 2015 programming foundations; 1–2 min in summaries of the American College of Sports Medicine's 2009 position stand) — justified at the time by an acute-hormone theory that the hypertrophy literature itself now questions, although the folk split of roughly 2–3 minutes for multi-joint lifts and about a minute for single-joint lifts sits closer to, and partly inside, the experimental 60–180 s plateau on which the standing 120 s setting rests. None of this makes the convention layer useless; it makes it useful for a different question. Self-report and convention measure preference, habit, adherence, and institutional fit — contaminated by placebo, cultural transmission, tradition, and the structure people already live inside, so that people largely report the schedule they inhabit. They can tell a practitioner which setting people will actually keep, which is a real constraint on any prescription. They cannot establish dose–response optima: a preference is not a performance curve, an observed average is not a peak, and a timetable slot is not a finding. Where convention and experiment conflict — short prescribed rests for hypertrophy, a 17-minute "best" break above the 1–10-minute experimentally tested band, 50- and 90-minute work blocks with no direct experimental counterpart at all — the experimental verdict governs the setting, and the convention layer explains, at most, why the setting will or will not be followed.