VIM concept: repeatability, reproducibility, and the conditions between

Precision is not one number. The VIM defines a ladder of measurement conditions, and the standard deviation you quote depends entirely on which rung you climbed.

Repeatability conditions (VIM 2.20) hold everything fixed — same procedure, operator, system, location, short time — and give the tightest, most flattering spread. Reproducibility conditions (2.24) swap laboratories, operators, and equipment, and give the honest between-lab figure that proficiency testing measures. The revised VIM's middle rung, intermediate precision (2.22–2.23), is the one most budgets actually need: same lab over weeks, with operators, standards, and days allowed to vary.

ISO 5725 built the classical statistical machinery for designed precision studies on these definitions. The everyday failure mode is quoting a repeatability standard deviation where an intermediate-precision one belongs — it understates the uncertainty by exactly the factors a working lab cannot hold constant. The dictionary carries all five condition-and-precision entries with their clauses.