X-charts and C-charts

[Steven_S_Prevette@rl.gov]

I would also like to clarify my position on the use of an X-chart in place
of a c-chart. 

Bottom line is that you can get about 95% of the functionality of the
c-chart by using an x-chart.  For people just starting out, if the x, p, c,
u series of charts is confusing, then fine, stick to x-chart.

There are two benefits of using the c-chart when measuring a Poisson process
- counting events:

1.  The spread of the data is compared to the Poisson standard deviation.
By comparing to the theoretical spread, we have the benefit of checking if
the data are indeed Poisson.  If the Poisson (c-chart) limits don't work
(the data are too narrow or too wide in their variation) then we have an
indication that the process is not Poisson, and this may indicate tampering.
Dr. Deming gives an example using a p-chart in OOC where he was able to
detect tampering by the inspector using the p-chart.  This tampering would
not have been detected on a x-chart.

2.  The calculation for the c-chart standard deviation is simple.  It is the
square root of the average.  I like using it when giving the Red Bead
Experiment because I can quickly calculate the control limits and make the
control chart of the Red Beads real time in front of the audience without
even using a calculator.   The square root of 10 is 3.2.  Yes, I know Dr.
Deming used the np chart, but the c-chart gives the same limits within an
error of a whole bead.

Steve Prevette
QA Engineer, ESH Radiological Compliance
Fluor Hanford, A Fluor Global Services Company
ASQ Certified Quality Engineer
steven_s_prevette@rl.gov
509-373-9371