ToolsGauge Bias Analysis

Gauge Bias Analysis

Evaluate systematic gauge bias. Bias = measurement mean − reference value; a one-sample t-test determines whether the bias is significant.

How to use: Measure a known reference standard multiple times, enter the measurements and reference value to determine if the gauge has systematic bias.

📥 Download Sample CSV

Measurements (repeat same reference standard)

Reference Value

Tolerance T (optional, for %Bias)

Use Case

When a new batch of calipers arrives, QA must confirm whether systematic bias is significant before deciding if calibration is required prior to use. An engineer measures a reference standard (known value) 25 times and uses a one-sample t-test to determine whether the measurement mean significantly deviates from the reference value. Gauge bias is a core MSA (Measurement System Analysis) item. AIAG recommends completing a bias study before commissioning any new gauge, confirming |Bias| < 10% of the measurement system specification width.

Example

Reference standard value = 25.000 mm Mean of 25 measurements = 25.042 mm Standard deviation s = 0.021 mm Bias = 25.042 - 25.000 = +0.042 mm (reads high) %Bias = Bias / tolerance width × 100% = 0.042 / 0.5 × 100% = 8.4% t statistic = Bias / (s/√n) = 0.042 / (0.021/5) = 10.0 p < 0.001 (bias significant) → Calibration required before production use

Further reading: Gauge bias analysis and calibration methods

View articles →

FAQ

Related Tools

Gage R&R Study

Bias analysis confirms accuracy; Gage R&R evaluates precision. Together they form the core of a complete MSA.

Open tool

Gauge Linearity Analysis

Bias analysis checks systematic error at a single reference value; linearity analysis checks whether bias varies systematically across the measurement range.

Open tool

Process Capability Index

Gauge bias skews CPK results high or low. Recalculate CPK after calibration to reflect true process capability.

Open tool