67824_en-US

liliconcom

ISO method for control charts for treating between-subgroup variability

I have a question about control limits in control charts when there is between-subgroup variability.

 

Have you heard of the following method for calculating control limits?

 

1. Calculate the traditional stddev, s , from all data values (with ignoring subgroup information).
2. Calculate sigma = s / sqrt (n) where n is sample size of each subgroup.
3. Set the control limits by [ m-3 * sigma , 3 + 3 * sigma ] where m is the overall mean.

 

This method is explained verbally in " 13.1 General caution "(pp.21-22) in ISO7870-2: 2013 (E) .
A user in Japan requests this "ISO" method, but I myself had never heard of it.
(I myself think that JMP's three-way control chart is more understandable than this "ISO" method.)

Re: ISO method for control charts for treating between-subgroup variability

I'm a bit unclear on what the standard is recommending? Does the standard recommend NOT using the formula you listed WHEN there is substantial between subgroup variation? While I've not read the standard, when you said it was a "General caution..." that makes me think this method is NOT recommended, which would lead to just the usual means by which JMP creates control limits for Xbar and R charts (little d2/R(bar) etc.)?

 

Of course the customer could also use the 3 way JMP approach for more information. Some consumers of the information contained in the charts might be confused though because quite frankly, I think most people are never taught how to use this 3 way method...they just go with the usual control chart calculations in Phase I and stop there.

 

I guess I'd agree with the recommendation...and the three way JMP approach is more informative.

Re: ISO method for control charts for treating between-subgroup variability

Thank you for your information and advice.

 

The standard is recommending the formula I listed when there is substantial between-subgroup variation. The standard says as below.

 

"On the other hand, Figure 8 shows another \ bar {X} and R chart for the same data as that in Figure 7, where the control limits of \ bar {X} chart are calculated on the basis of the overall process variability instead of the mean of ranges (\ bar {R}). "(p.21 in ISO 7870-2: 2013 (E) )

 

This explanation is not clear, but the user has a data set to reproduce the LCL and UCL which are shown in the standad as an numerical example. 

 

JMP's or traditional control limits in Xbar-R charts and Xbar-S charts are calculated only from within-subgroup variation. The method I listed calculates the standard deviation from all data values. This standard deviation can be calculated by the Distribution platform without setting any By-variables.

Re: ISO method for control charts for treating between-subgroup variability

The purpose of rational subgrouping is to minimize the variation within a subgroup and monitor the variation between subgroups. By calculating the standard deviation, they are ignoring the two sources of variability. Why even subgroup the data? Why plot the mean?

 

As my colleague Di Michelson said:

"If there is different within and between variability, you should acknowledge that and use the appropriate variance component on the control chart. It’s easy to use the “right” formula for the variability of the plotting statistic, no need to get fancy with “wrong” formulas."

 

If they want to ignore the different variability sources, it seems to me that a Levey-Jennings chart (which uses the standard deviation) on individual data points would be more appropriate. Just not sure what this approach is trying to accomplish.