Does anyone use Median Charts instead of Xbar&R charts?

My dream, but I tink that most of practitioners feel uncomfortable with non parametrics.

Althought Wheeler showed that the use of traditional SPC worked fine with other distributions (for rule #1 >> out of specs), the application of other rules sometimes has problems (especially with continuos data, ie. chemical batch). I even tested rules with a random number generator and always happen.

:nopity:

The question is why to use Median SPC??

If there is to many out of control rules, you may need to take some out, not all rules apply to all (there is no one size fit all).

If you want to represent more exactly the process, go a head (no problema).

If the control limits always show out of control limits, and there is expected some degree of autocorrelation in your data, affect the limits with the autocorrelation coefficient.

I tink the question it self is too general, and Median SPC may be known but not too often used.

:agree1:

I have been playing around with this template. Not sure if this helps you but you're welcome to it.

Does anyone use Median Charts instead of Xbar&R charts?
Davis Ballestracci, the current chair of the Stats SIG at ASQ, is a big proponent of median SPC. He has a book out called "Quality Improvement Practical Applications for Medical Group Practice". There are some advantages, such as it is very impervious to outliers, but there are losses in statistical power.