Hello all!
Greetings. I am involved in developing SQC tools. Recently a client
asked whether it is possible to design a EWMA or CUSUM type control
chart for the process <variance> (or <standard deviation>). My first
reaction was that it probably would not be a good idea, since the
variance has a probability distribution that is not Normal, or
additive. It seems, though, that CUSUM/EWMA Control Charts for the
variance have been developed, e.g., A New S2-EWMA Control Chart for
Monitoring the Process Variance, Philippe Castagliola, EWMA for
monitoring a process standard deviation, Crowder & Hamilton. I have
just downloaded the first paper (http://www3.interscience.wiley.com/
user/accessdenied?ID=110494448&Act=2138&Code=4719&Page=/cgi-bin/
fulltext/110494448/PDFSTART), which is apparently an improvement over
the earlier Crowder & Hamilton paper (Crowder SV, Hamilton MD. An EWMA
for monitoring a process standard deviation. Journal of Quality
Technology
1992; 24(1):12-21)- I have not seen this paper. the basic idea of
these authors seems to be to use a logarithmic transformation to
"normalise" the variance.
So now (before I try to program and make this tool available to my
client!) my question is: does anyone here have experience of using
these charts? I don't think they have been incor****ated in any SPC
software that I have come across, which adds to my reservation!
Thanks and Regards
ESG
|
