I wanted to cite the reference for the %contribution. GRR=GRR9. AIAG only provided the guidelines on %StudVar and ndc. Who first recommended the limits for %contribution? One which I could cite.
Dear Prof Abu
Only available reference for Citation is AIAG... but many material on Six Sigma Training materials etc, cite less than 10% is the target, 10 to 30% is OK and above 30% is not accetable...
As the Special cause variations comes from 1) Part to part 2) Machine & Process 3) People and 4) Instrumentation, it is a good idea to keep variation contribution from the Instruments at the lowest level.
With extensive experience in GE and specialised in GRR, RAISE specilaises in training programs with clients like ACMA (Automotive component Manufacturers in India) and corporate houses.
Kindest Regards
Ramanan
Thanks Ramanan,
Some like me, have been made aware of the mathematical flub of using the non-additive standard deviation based metrics to determine %Std Var (in percentage) as proposed by AIAG. Hence we will like to confine ourselves to the variance based metrics i.e %contributions. I am also made to understand that 9% is estimated directly from %Std Var, (the value 0.01 and 0.09 was derived from 0.1^2 and 0.3^2). However this was never explicitly stated in AIAG guideline, hence citing AIAG would be improper. I am also made to aware that six sigma do use the %contributions limits in their training.
But those figures can't have just dropped from the sky! Someone must have thought about it and introduced it. But who? When we square 0.3 we not only square the std deviation into variance but also the k value within (either 5.5 or 6.0)
Has this step been validated; (6*sqrt(var))^2? %contribution is estimated by direct ratio of variance component unlike that of %stdvar which is std dev multiplied by k . Why then should %contribution benchmark include the k value, with the k value now bloated by squaring, to 36!
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS