Gage R&R (Repeatability and Reproducibility) Destructive Test Method

B

Benoit

I have realized a R&R study on destructive method (measure of moisture on plastic granules). I've taked ten sample from different production lots. I have split this sample in two parts and I keep this samples in the same conditions. The test were realized by two appraisers in a short time.
I Have used the range method from MSA Third Edition Page 97.

It is correct ?


:confused:
 
R

Ravi Khare

It is preferred that the destructive test R&R analysis is done using the ANOVA method.You should use Nested ANOVA, since every sample is not replicated across all operators.
 
D

Darius

"I have split this sample in two parts and I keep this samples in the same conditions. "

I hope, you are spliting each sample in two parts and that in the sampling process, you take in account the lower and the upper conditions of the process (increase the variation between subgroups = that the moisture production lots represent all the posible variation of the process). Most of the people I know, don't care about it, but it's the most important point (if you like to see R&R differences across appraisers)

The latest version of Minitab (I don't like advertaising, but you can download a trial) has Nested ANOVA, but I readed the description, and looks like is just to take in account that the variation between consecutive samples are almost nil (I was very disapointed ).:bonk:

There is almost any information about destructive testing, my recomendation is just to take rational sampling in account and maybe, split each sample in more subsamples (more trials may show the real appraiser variation).
:frust:
 
Top Bottom