Whenever we collect data for a project, we need to make sure the measurement system we’re using gives good results. We want to ensure we have an effective measurement system for the characteristic that defines the problem. Nothing is worse than conducting an analysis on faulty data because the measurement system was not adequate.
The measurement system consists of the gages, people, methods, material, and environment involved in measuring a characteristic. This system provides a window to view the process. If the window is foggy, we are not able to see clearly what is going on.
There are several important attributes of the measurement system that need to be evaluated to determine its effectiveness:
- Measurement variation, also called precision, is the standard deviation of the measurement errors.
- Measurement bias, also called accuracy, is the average measurement error.
- Discrimination, or the ability to see differences in parts.
These and other measurement characteristics are explained thoroughly in the AIAG, Measurement System Analysis booklet.