Choose Language Hide Translation Bar
Highlighted
OneNorthJMP
Level IV

XBAR R & S don't display control limit?

Why XBAR R & S don't display proper control limit ? 

 

image.png

4 REPLIES 4
Highlighted

Re: XBAR R & S don't display control limit?

You have varying subgroup sizes.  This means you will also have limits that vary rather than constant limits.  To see the limits for each subgroup, select Save Summaries from the red triangle menu.

Highlighted
OneNorthJMP
Level IV

Re: XBAR R & S don't display control limit?

I can see that save summaries will give each lots different control limit. But that is not what i want, JMP should able to provide control limit by subgroup. Since i have 2 subgroup, there should be 2 subgroup control limit.
Highlighted

Re: XBAR R & S don't display control limit?

From your image, there are more than 2 subgroup sizes.  Phase RB79 has at least 2 different subgroup sizes.  SN100B2 also has at least 2 different subgroup sizes.  So there would have to be at least 4 different sets of limits displayed for this example.  You can imagine that if every subgroup had different sample sizes for your example, there would be 13 sets of limits.  This could make the Limits Summary report unwieldy for large data sets.

Highlighted
statman
Level VII

Re: XBAR R & S don't display control limit?

The bigger issue is not what JMP will display, but the fact that you are changing subgroup sizes in side your study.  This is very dangerous since the factors that make up the within subgroup variation are changing in your study.  The purpose of the S chart is to determine if the within subgroup variation (which is a function of the variables changing within subgroup) is consistent or stable.  If that variation is unstable, then you should seek to understand why.  If that variation is stable, then the sources of variation within subgroup may be compared to the sources of variation between subgroup (due to the variables changing between subgroup) to determine which source has greater influence on the variable being charted.

 

If you change what is captured within subgroup (by changing the subgroup size), you defeat the purpose of the control chart.  The comparisons are changing in your study.  For further information on appropriate use of control charts, you should review Shewhart's "Economic Control of Quality of Manufactured Product". or Wheeler's book "Statistical Process Control".

Article Labels