Number of Distinct Categories" or NDC Calculation in MSA Studies

Jim Wynne

Leader
Admin
Deming knew there were issues that were unknowable. This isn't one of them.
The fact that some quantity is knowable doesn't necessarily mean that it needs to be known. I've quoted Bob Dylan many times here over the years: You don't need a weatherman to know which way the wind blows. Of course this isn't true in all cases. There are several different skills needed in order to be able to make judgement calls like this, and one of them is understanding the system/process well enough to know where the line is between prudence and overkill. If we constantly allow the tail to wag the dog, there will be much wasted time and effort. Another one I've used here many times came from Peter Drucker: There is nothing so useless as doing efficiently that which shouldn't be done at all.
 

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
I suppose one could say you should not bother to do a lot of quality approaches if you don't know how to do them right...you ARE wasting your time because doing it wrong IS useless. But, that is not the approach's fault. There is no point in measuring if the measurement is inadequate. If there was no need for adequate measurement, just invest in yardsticks, rulers, protractors and walk away. Keep it simple...third grade simple.
 

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
The youden plot for standard R&R is great. the multi-vari (Seder) is perfect for most NDE, destruct, and other weird systems. both plot the data that you need to see in a form that is simple and straightforward - no mental gymnastics required.
I would love to see some examples!
 

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
a couple of examples are in the attached presentation from the 2019 World ASQ conference.
Excellent! Very good starting point. Looked like it had a dose of Wheeler's issues with Gage R&R in it, also.
 

Bev D

Heretical Statistician
Leader
Super Moderator
I don't think of the concerns that Wheeler has documented are mere issues. (he isn't the only statistician either by the way)
The facts are that the popular method is anchored by fake math. and the 'goals' for determining the goodness of the R&R are not theoretically or empirically derived. they were pulled form thin air.
As professional quality practitioners, engineers and frankly adults we shouldn't support or rationalize fake math or arbitrary 'goals'. soon there will be no science or math or truth...
 

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
Dr. Wheeler and I joked about how AIAG would quote him, but not listen to him. But, it is no different when fake math supports Shewhart charts for all applications...well, more like urban legend. Although many statistical approaches do require a lot thought-I am still not opposed to bright lines when they work. Even well lit lines are better than totally rolling over to Quality Level TCE. In any event, certainly going to review the approach to see if it answers the questions I expect to be answered when doing gage analysis. You know I have no fear of that. Will be cool if it does.
 

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
...soon there will be no science or math or truth...
Well, after looking at the treatment of covid data by people using statistics (cannot call them statisticians) we are pretty far away from science, math or truth now more often than not.
 

Ninja

Looking for Reality
Trusted Information Resource
Off topic, but we're headed there already...
In grad school, I put a book on my shelf, title cover facing the door, which read "Statistical Manipulation of Data. How to get data to tell you what you want to hear."
Advisors were not impressed...but both borrowed it at least once...
 
Top Bottom