In his February 2017 Quality Digest column, “Don’t We Need Good Measurements?” Donald J. Wheeler recommends that a measurement system contributing up to 80 percent of the overall variation (on the variance scale) is good enough to detect persistent mean shifts when using a process behavior (control) chart. As a result, he concludes that assessing the quality of the measurement system before implementing the chart is likely a waste of resources and time.
ADVERTISEMENT |
We disagree with both his argument and conclusion. We suggest that you first look at Wheeler’s December 2010 column, “The Intraclass Correlation Coefficient,” a reference he kindly provided us. This column describes the intra-class correlation and provides additional details about the example discussed in the 2017 article.
Wheeler uses the model
…
Comments
Boldness
Right or wrong, congratulations on your boldness in attacking the world's leading expert on quality. Too often people are afraid to speak up. Only by stepping forward into open discussion, can quality progress. Very much looking forward to Dr Wheeler's enlightening reply. No matter who proves correct, we will all learn.
Different interpretation
Interesting article. I enjoy reading alternative viewpoints on this subject.
My interpretation of Wheeler's discussion of intraclass correlation and measurement systems differs. I understood his argument as prioritizing process behavior charts over measurement system analyses. He demonstrated that a poor - but stable - measurement system is still capable of detecting "signals of exceptional variation" provided the charts are constructed and interpreted properly. Wheeler goes on to imply that the absence of process signals should prompt the investigation of the measurement system.
To me, the implicit point of Wheeler's discussion was that in the reality of limited time and money, one is better off charting the process measurement data to identify the "signals of exceptional variation" so they can be dealt with. The process behavior charts are robust enough to overcome measurement error in the presence of exceptional variation. Once the process is stable (homogeneous), a deeper investigation of the measurement system may be in order.
Good Summary
BILLS, I could not have said it better or more consicely.
Thank you!
Add new comment