Re: My Sigma's bigger than your's

[GrantBlair@aol.com]

In a message dated 10/8/2000 7:17:59 AM Eastern Daylight Time, 
jsdwd@samart.co.th writes:

<< Would it be your position Mr. Blair that the new fixture should not be
 introduced to the process until all the other special causes were
 eliminated? >>
There are two obvious possibilites for error:
1. Both fixtures may have been tested when the process is stable, but the
    old fixture may have simply averaged higher, the new fixture lower. When
    one averages two sets of data, this is generally the case.
2. One fixture may have been tested when the process had a special cause,
    the other during a stable period. I recall watching a statistician defend 
a higher-   order interaction found in a DOE and having to struggle to 
explain the lack of     main effects until it became apparent he had made one 
run at the same time a      special cause occurred
The remainder of your post reminded me of another incident from my 35 years 
of experience. (which means my sigma is usually smaller than yours. I may not
always like it, but that's the way it's supposed to work };-)
I was interrupted in the middle of a presentation to staff on a new test 
method by a respected engineer who angrily declared that the relationship I 
was showing was
backwards from the true relationship. The rest of the staff joined in for 
what became a heated discussion. I left the meeting and returned to my office 
with the quality director barking at my heels, urging me to compromise my 
position.
About 15 minutes later, I received a apologetic call from the engineer. Upon 
returning to his office, he had reviewed his data, and realized I had 
correctly stated the relationship. Since my manager was there (still chewing) 
, I simply handed him the phone.
Later, the manager asked how I could remain so calm throughout all this. I 
replied,
"Because I knew God hadn't changed any of His Laws when I got up this morning.
Regardless of how many people were against me, I knew God was on my side"

Grantblair@aol.com
Ninety Six SC
========================================================================