Fatigue is an integral; a short trial never sees it

What it is

The cost of a raised arm is an integral. A thirty-second usability trial, three successful waves at a booth, stop in the region where the integral has not yet become ache and tremor. People in a short trial may even report “pretty easy,” because metabolites have not reached the complaint threshold. Taking that comfort score as a voucher for a product that must last a film, a class, or a surgical case is measuring with the wrong time window. Accumulation also has recovery: putting the arm down for thirty seconds versus three seconds changes the start of the next bout. If the test does not include long enough work and real pauses, the second half of the curve stays unseen.

Why it happens

Local muscle fatigue is not a switch. Metabolism, ions, and motor-unit recruitment drift in time; the recognizer sees growing jitter and shrinking amplitude. Short trials sample before the drift. Rating scales are sensitive to “what I just did,” not to “what this will be in ten minutes.” Practice can raise recognition in a short window while fatigue lowers it in a longer one—the two time scales point opposite ways. Reporting only the short window yields the illusion that it gets better with use. Pause structure is part of the mechanism: in a demo, the presenter's talking gaps are used to drop the arm; live continuous use has no such gaps.

Studying it

Protocols must declare duration and pauses, and must reach the same order of magnitude as the claimed use (a whole episode if you claim a film, not the trailer). Repeat measures: jitter, amplitude, recognition, ache at start, middle, end. Allow real pauses (watching content, talking) and log arm drops. A short trial as a control shows how it misses the second half. Recovery curves: drop for different durations and see whether recognition and ache return to baseline. Do not take a single comfort score while people are still excited from having just learned the move.

Where it stops holding

An interaction that only ever lasts tens of seconds (a gesture to pass a door) does not need a long protocol; the short trial is the right window. Game levels with rest screens have pauses designed in; the continuous combat stretch still needs the long window. Laboratory “hold until failure” measures maximum endurance, not accumulation in real use, and will overstate pain and understate stealth unloading. Drugs, sleep, and yesterday's training change baseline; long studies need to control or at least log them.

Applying it

• Put the claimed continuous-use duration in the test plan. Comfort data shorter than that duration may not stand alone as ship evidence.
• Log jitter and arm drops at start, middle, and end. If the second half worsens, change posture or modality; do not add a short teaching clip.
• Label booth demos “fatigue window not included,” so sales cannot replace a ten-minute task with a thirty-second success.

Related

• Same group: C4.12.1 An unsupported raised arm fatigues in minutes · C4.12.2 High-frequency actions must not demand overhead or a long hover · C4.12.4 Long tasks need a supported posture or another modality
• Adjacent: C4.05 Three dimensions of gesture evaluation · C4.09 Finger count as input
• Search: fatigue accumulation · short trial bias · endurance protocol