Use a gage linearity & bias study to assess linearity (how accurate your measurements are) and bias (how well your measurements compare to a reference value).
Answers the questions:
- Does my gage have the same accuracy for all sizes of the objects being measured?
- How biased is my gage when compared to a master value?
|When to Use
|Start of project
||To verify the accuracy of a gage over the expected range of outputs anticipated at the start of the project to ensure that accuracy is consistent prior to performing a baseline analysis.
|Start of project
||Verify improvements did not affect the accuracy of the measurement device. For example, the improvement may have resulted in producing outputs that are consistently larger or smaller than what was typical before the project.
Your data must be a continuous Y with known standards available.
- Samples with known reference or master values must be available over the full range of expected output and gage output must be continuous.
- If you have discrete numeric data from which you can obtain every equally spaced value and you have measured at least 10 possible values, you can evaluate these data as if they are continuous.
- Select parts with known exact values. Be sure to collect parts covering the full range of normal output.
- Measure each part multiple times using the same appraiser with the same gage.
- Enter the part identifier into one column, the actual readings into a second column, and the reference values into a third column.
- Optionally, enter the process variation (6 x StDev) in the dialog box. This process variation is used to calculate the % linearity and % bias relative to the width of the process (±3σ). You can get this number from your gage R&R study.
For more information, go to Insert an analysis capture tool.