Re: Tom Ryan's reply to SPC Theory

[Kromkowski@aol.com]

>>When we use the latter we naturally use 3-sigma limits because we assume 
that
we have a symmetric distribution, and ideally the distribution should be 
close to a normal distribution. If not, the chart will have properties much 
different from what is assumed.
We should generally not use 3-sigma limits for attribute data because
the appropriate distributions are skewed, and may be greatly skewed.<<

In my view, the underlying distribution is usually insignificant to the use 
of the Ind.X/Mr chart. Hence, using 3 sigma (where sigma is approximated by 
Shewhart's method of R-bar/d2) is entirely appropriate. What disturbs me the 
most about the post, however, is for you to write that 3 sigma's shouldn't be 
used without providing a reasonable, understandable, and economical choice 
for how many sigma should be used.  

It need not be rocket science -- learning when to use a p, c/u, or x flavor 
chart is not difficult, and when in doubt a Indx/Mr in time order or in 
random order is not going to lead one too astray in the task of discernment 
between likely special causes and common causes of variation.

JDKromkowski
kromkowski@aol.com