What data do you need?
Prepare part identifiers, operator identifiers and repeated numeric measurements. Choose parts that represent the process range, randomize measurement order and use a consistent procedure. A crossed study measures the same parts with each operator.
Worked example
A study with 10 parts, 3 operators and 2 repeats per part/operator combination produces 60 measurements. This is a planning example, not a mandatory design. Record every measurement with its part, operator and repeat identifiers.
How to use this Six Sigma tool
- Select representative parts and plan operators and repeats.
- Collect measurements in randomized order and open Gage R&R Studies.
- Review measurement variation before establishing a process baseline.
How to interpret the results
A measurement system can obscure real process differences. Review the study design, variance components and the denominator used for percentage results. Percent of study variation and percent of tolerance answer different questions; use the criteria required by your organization.
Should I run Gage R&R before process capability analysis?
Evaluate measurement quality before relying on capability results. If measurement variation is substantial, observed process variation may reflect the measurement system as well as the process.