4. Linearization of standard deviations

Dispersions are often influenced by the "quantity" measured. For example, it's easier to measure a 10 mm mechanical part than a 2,000 mm one. The standard deviations of repeatability and reproducibility (inter-something or overall) can therefore differ according to the quantity measured.

To confirm or refute this hypothesis, it is possible to carry out tests by varying the quantity measured. We're talking about different levels here.