Using data and control charts in management

[SAdamsTimD@aol.com]

We have found that the XmR chart on higher level issues are also useful.   
Many times when one looks at numbers one is tempted to ask "what happened?" when 
the better question is " did something happen?" 

If the control chart indicates that an out of control event has occurred, 
then it is appropriate to search for a cause, even if it is outside this system 
to possibly explain the change. If the numbers are unsatisfactory and the chart 
does not indcate an out of control condition, then one must do the difficult 
job of really understanding the system that generated the data and coming up 
with a system improvement theory.

I have observed that so many systems are in control that I submit that to say 
"who is at fault?" is highly suspect; "what (sysstem) is at faul?" t is a 
better theory almost all of the time.

If profits went down this month or accidents went up, one can jump to the 
conclusion that someone did something wrong when it is probably just normal 
variation just as if ER cases went down the month that the ship went out, we may 
figure that these two events are casual when a control chart will prove that one 
should even start look for casuality.

I recommend Dr. Don Wheeler's book, "Understanding Variation" for a good 
lesson is these types of understandings.

Stan Adams
The Time Dimension of Quality
Pensacola, Fl