A Gage R&R study quantifies how much of the variation you are looking at comes from the measurement system rather than from the thing being measured. Repeatability is the variation when one person measures the same item repeatedly; reproducibility is the variation between different people measuring it.
Every conclusion drawn in Measure and Analyse rests on the data being real. If a third of the variation under analysis belongs to the gauge, the analysis is describing the measurement system and the resulting improvements will be aimed at nothing.
It is the most commonly skipped step in Six Sigma projects and one of the most reliable reasons a technically sound project produces no result. It also takes a few hours, which makes skipping it hard to justify.
The usual thresholds: under 10% of total variation is acceptable, 10–30% depends on what the measurement is being used for, above 30% means the measurement system is the project until it is fixed.
Studies get run on parts that do not span the real range, which flatters the result. And a poor result gets read as a reason to abandon the project when it is simply a finding — a bad gauge is something to fix, not a dead end.
