For GR&R (Gage R&R), is multiple testing of the same sample necessary?

[Browncoatb]

Hi, i'm trying to do some GR+R on two tests we currently perform, one on mon/wed/fri and the other on tue/thur. My question arrises because for the tue/thur test, half of the actual test takes 3 1/2 to 4 hours to complete. Im not sure because im an engineering intern, and this is the first time i've actually used any of the statistics i learned in class,

thank you much

 

[Jim Wynne]

Re: For GR&R, is multiple testing of the same sample necessary?

Hi, i'm trying to do some GR+R on two tests we currently perform, one on mon/wed/fri and the other on tue/thur. My question arrises because for the tue/thur test, half of the actual test takes 3 1/2 to 4 hours to complete. Im not sure because im an engineering intern, and this is the first time i've actually used any of the statistics i learned in class,

thank you much

Welcome to the Cove.

What kind of testing is this? Why do you feel that GR&R (as opposed to some other form of analysis) is necessary or might be helpful?

 

[Browncoatb]

Re: For GR&R, is multiple testing of the same sample necessary?

I'm actually not sure at all, i need to find a way to see how effective, and/or accurate our current testing procedure is, and in passing my boss mentioned the GR+R so im looking into that first. The only problem is that currently, we dont perform any multiples of tests, only one test on one sample taken either M/W/F or T/Th

 

[Jim Wynne]

Re: For GR&R, is multiple testing of the same sample necessary?

I'm actually not sure at all, i need to find a way to see how effective, and/or accurate our current testing procedure is, and in passing my boss mentioned the GR+R so im looking into that first. The only problem is that currently, we dont perform any multiples of tests, only one test on one sample taken either M/W/F or T/Th

What type of testing? What's the manufacturing process?

 

[Bev D]

Re: For GR&R, is multiple testing of the same sample necessary?

can you test the thing more than once?

 

[Browncoatb]

Re: For GR&R, is multiple testing of the same sample necessary?

Its the papermaking process, and specifically we are determining cloride residuals on white liquor (the M/W/F test) and sulfidity content on the smelt from the recovery boilers (the Tu/Th test). the white liqour is the chemical wash supplied to the digester which turns wood chips into a sort of dissolved-wood slurry, or pulp, that then gets made into paper, and the smelt is the molten chemicals that remain after the spent, black liquor is burnt for heating value. and yes, it is possible to take the tests multiple times, im just not sure if i should be proposing a change in the the testing procedures when i dont even know if the current procedures are effective.

 

[Bev D]

MSA (commonly called GR&R) is intended to determien the effectiveness of any test measure.

The general answer - with only one exception - is that YES you must test each sample twice to determine the repeatability of the test. Repeatability being the most fundamental assessment of effectiveness.

In the case of chemical batches liek yours, I have found that the best approach is to take a sample that is large enough to be split into multiple test samples. homogenous the batch (such as by stirring) to minimize any within batch actual variation.

thsi testing protocol is only done to assess the R&R of the system. If the system is adequate you only have to test teh sample once in normal production.

 

[Browncoatb]

excellent, thank you very much. that was my biggest question, that of "how can the MSA be effectively telling us about out current testing procedure if we have to change our testing procedure in order to properly analyse?"