attribute gage RR in a unique way

[dsnell]

I need to perform a gage R&R on a test that we do which checks for blockage in very small tubes. the problem is that we hold them in bundles as large as 300 pcs. my plan is to have typical bundles and have 25-30 pieces knowingly blocked and ensure the inspectors can detect them. my question is how would I explain that in minitab when performing the analysis? would I simply list out 300 lines and have most say accept except for the failures? and how would I know that operators were in agreement of which samples were rejected?

 

[Ninja]

Is it safe to assume that if one tube has a blockage, you fail all 300?

If so, the test results are actually numbers from 0 - 300 with a spec maximum of zero?

 

[dsnell]

We would not treat one bad as all 300. We simply pick out the bad ones. And throw out any blocked.

 

[PaulJSmith]

I think one thing you need to remember about R&R is that it's not about ensuring that your inspectors can succeed, but rather that it's about finding out the margin of error in their success/failure with that particular gage.

 

[Bev D]

yes you should put 300 lines in minitab with their pass fail status.

number each tube some way so that you can have traceability back to what each operator rejected or passed and whether or not the tube was a pass or a fail...