Calculating instrument dispersion
To calculate the dispersion of a measurement system, the repeatability and reproducibility test is used.
Figure: Repeatability and Reproducibility
The repeatability and reproducibility test aims to characterize the entirety of the dispersion of a measurement system by separating that which is the responsibility of repeatability (dispersion of measurement repetition) and that which is the responsibility of reproducibility (the difference between several operators).
In order to carry out this test the following is necessary:
Standard test: 3 operators each measuring 10 parts, 3 times per part
Rapid test: 3 operators each measuring 10 parts 1 time per part (this test does not allow for the disassociation of repeatability and reproducibility)
Once the measurements are taken there are two methods for calculating the dispersion of the average measurement.
RANGE method: This method is more approximate and easy to calculate with an Excel worksheet. However, it does not allow for the detection of possible interactions between parts/operators.
The ANAVAR is recommended as it is the most precise.