Endurance.Net Ridecamp Archives

Sisu West Ranch wrote:
> Disclaimer: Do not interpret this post as a distrust of statistics and 
> statistical interpretation of data.  Repeat:  WHILE I AM NOT AS VERSED 
> IN STATISTICS AS TRUMAN, I USED THEM CONSISTENTLY WHILE EMPLOYED AS A 
> RESEARCH CHEMIST.
>
> What I noticed was that in the real world of manufacturing products 
> even when the usual statistical tests showed significance to a very 
> good degree of certainty it was not unusual to find that over the 
> longer term, the trend reversed if the difference was small.  Also, 
> even when the effect was real, but slight the usefulness of the 
> information was also slight.
However, that is what quality control in industrial processes is all 
about. When sampling indicates a "change" that is resulting in a 
decrease in quality - number of defective products compared to the total 
number of products for example is detected - then management wants to 
know since it represents real money. If for example an incoming shipment 
of steel bar stock used by Whirlpool corp to make rotors for 
refrigerator compressors all of a sudden jumps by a small amount in the 
number that fail the incoming test - both physical and chemical analysis 
the whole lot is rejected as being out of spec. That is a true example, 
by the way.

If your defective rate is nominally one in 1000 and it goes to 4 in 1000 
(.3% difference) that can add up to a lot of additional cost to a 
company. While very few cars for example get recalled compared to the 
numbers produced - a recall or even replacing defective parts on 
warranty can be extremely expensive.

Threshold for change detection in quality control are normally higher 
than what was used in the quick calculation of pulls.

Hence in reality depending on what you are testing for - dictates the 
how small a difference you will accept as your threshold.

What is the practical  use of knowing a HW is more likely to get pulled 
- in general mathematicians don't normally bother themselves worrying 
about practical use :-) ! 

Truman

--

"The trouble with fighting for human freedom is that one spends most of 
one's time defending scoundrels. For it is against scoundrels that 
oppressive laws are first aimed, and oppression must be stopped at the 
beginning if it is to be stopped at all."   H. L. Mencken


=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

Ridecamp is a service of Endurance Net, http://www.endurance.net.
Information, Policy, Disclaimer: http://www.endurance.net/Ridecamp
Subscribe/Unsubscribe http://www.endurance.net/ridecamp/logon.asp

Ride Long and Ride Safe!!

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=