Definition
Assess the process of measuring data quality and quantifying the variability of measurement systems.
In detail
Gage Repeatability and Reproducibility (GRR) is a method for evaluating measurement system variation to determine whether measurement data is trustworthy. It addresses the question of whether measurement error masks true changes during process improvement. Compared to Measurement System Analysis (MSA), GRR is a core component of MSA, which also includes bias, linearity, and other analyses. GRR typically outputs the percentage of variation (%GRR) to judge whether the measurement system is acceptable.
Example in practice
In a quality improvement project, the team needs to measure product dimensions to determine process stability. They first conduct a GRR study, having three operators measure the same batch of parts, calculating repeatability and reproducibility. The results show %GRR is 25%, which falls in the 10%–30% "evaluate before use" band: given the measurement will be used for process improvement, the team judges it not reliable enough, so they improve the measurement tools and train the operators, then re-run the GRR to confirm the data is reliable before further analysis.