Evaluation of Measurement Uncertainty

Overview of Type A and Type B Evaluation of Uncertainties

The analysis of a measurement system involves understanding the uncertainties associated with taking a particular measurement and then, where possible, quantifying those uncertainties. Evaluation of measurement uncertainties are classified into Type A and Type B.

Type A Uncertainties can be quantified by statistical means. Type A Uncertainties include issues with:

• Accuracy
• Linearity
• Stability
• Repeatability
• Reproducibility

Accuracy

• Implementation and use of a solid Calibration Program helps address accuracy issues.

Linearity

• Looks at the accuracy of measurements over the full range of the device and can be evaluated using statistical methods.

Stability

• Stability studies evaluate a measurement device performs over time.

Studies of repeatability and reproducibility are called Gage Repeatability and Reproducibility studies or GR&R for short.

• The variation from the test equipment is called Repeatability. A good way to think about repeatability is that it is the variation we get when we measure the same part a number of times with the same measuring device.
• Reproducibility variation is the uncertainty, or variation, we get between people who are measuring the same parts with the same gage. Many times this is the largest source of measurement system variation.

There are other uncertainties associated with measurement systems that cannot be evaluated as readily by statistical methods. These are called Type B Uncertainties.

• A Type B Evaluation of Uncertainty is typically based on engineering and scientific judgment and not on statistical methods.
• Type B Uncertainties include issues related to:

• Temperature
• Errors in part form
• Fixture variations
• Conditioning requirements