gage reproducibility

Gage Reproducibility is the variation in measurements when DIFFERENT operators (or different conditions) measure the SAME part with the SAME gage — also called appraiser variation (AV), it measures the additional measurement uncertainty introduced by different operators, shifts, or setups.

Reproducibility Assessment

• Method: Multiple operators each measure the same set of parts multiple times — ANOVA separates operator variation from repeatability.
• AV: $AV = \sqrt{(\bar{x}_{diff} / d_2)^2 - (EV^2 / nr)}$ — appraiser variation corrected for repeatability.
• Sources: Operator technique, measurement setup differences, recipe interpretation, sample alignment.
• Automated: In highly automated semiconductor metrology, reproducibility may also test different tools or recipes.

Why It Matters

• Training: High reproducibility variation indicates operators need better training or procedures need standardization.
• Automation: Semiconductor fabs minimize operator reproducibility by automating measurement — recipe-driven measurements.
• Tool-to-Tool: In semiconductor, "reproducibility" often means tool-to-tool matching — different CD-SEMs measuring the same wafer.

Gage Reproducibility is person-to-person consistency — measuring how much variation different operators or conditions add to the measurement result.