Re: SPC on the rocks

[John <jsdwd@ispwest.com>]

On 7/9/03 4:01 PM, "Jean-Marie Gogue" <jme.gogue@wanadoo.fr> wrote:

> Deming was definite. No condition is required for drawing a control chart
> with experimental data and attempting to understand if there are special
> causes of variation or not.
> 

I agree with Jean-Marie.  There are no pre-conditions to drawing a
time-sequence chart (run chart) as long as one has time-ordered data.
Putting limits on such a chart is always a bit of a guessing game, even when
the so-called 'necessary assumptions' have been met.

I think it's helpful to remember that the state of statistical control is
only operationally defined.  Who knows if there is such a state.....

With regard to independence, I'm going to be lazy and not dig up a bunch of
references, but hasn't Wheeler sown that a fair amount of auto-correlation
(clear lack of independence) can be encountered without materially affecting
the usefulness of the control chart?  Perhaps some additional comments along
this line would be helpful to those who are confused.

John Dowd
jsdwd@ispwest.com