Re: SPC Theory Question

[Kromkowski@aol.com]

Although, I have never read either Chambers or Wheeler, I have also come to 
the same conclusion intuitively, which some empirical checking including by 
way of simulations.  My experience is that Jean Marie Gogue has also seem to 
come to the same conclusion, of course he can speak for himself.

I'd go so far as to say (although, I suppose Wheeler and Gogue might 
disagree) that even when the data is not time ordered, the IndX-MovR chart 
can be used so long as the data is ordered randomly (which is easy enough to 
do with computers these days) or some means that is not relevant (for 
example, alphabetical).  This is because the chart is just a tool for 
estimating three stds in a way that tends to account the messiness and 
relatedness of real life data better than calculating sigma by using the 
square root of the variance where samples are small no matter the 
hypothesized idealized distribution.

John David Kromkowski
Kromkowski@aol.com