bias

**Bias** in metrology is the **systematic difference between the average measured value and the true (reference) value** — a constant offset that affects accuracy (not precision), caused by calibration errors, measurement physics, or systematic instrument offsets. **Bias Assessment** - **Reference Standard**: Measure a certified reference material (CRM) or NIST-traceable standard — compare the average measurement to the certified value. - **Calculation**: $Bias = ar{x}_{measured} - x_{reference}$ — positive bias means the gage reads high. - **Significance**: Perform a t-test to determine if the bias is statistically significant — small biases may be within noise. - **Correction**: Apply a bias correction: $x_{corrected} = x_{measured} - Bias$ — calibration removes systematic bias. **Why It Matters** - **Accuracy**: Bias is the primary component of measurement accuracy — precision (repeatability) and accuracy (bias) are independent. - **Calibration**: Regular calibration corrects for drift in bias — calibration intervals must prevent excessive bias accumulation. - **Tool Matching**: Bias differences between tools (CD-SEM #1 vs. #2) cause apparent process variation — matching requires bias alignment. **Bias** is **the systematic error** — the constant offset between what the measurement tool reports and the true value, correctable through calibration.