Gage R&R Study by Attribute (ACCEPT-REJECT-INVESTIGATE)

J

Jeenan

Hi Everyone, I'm trying to do a study by attribute.....here are the details:

- We had 25 parts to categorize
- We had 3 operators
- Each operator checked the parts (25) 3 times
- We had 3 possibility as result (ACCEPT - REJECT - TO BE INVESTIGATE WITH A MORE PRECISE GAUGE)


We try to analyze the % of repetability on a single operator & the % of repetability (agreement) between the 3 operators

I would like to add that we KNOW the actual condition of the part, so it would be great to analyse if the decision (ACCEPT-REJECT-INVESTIGATE) of the operators were the good one.....

Thanks!!:thanx:

Jean-Pierre :agree1:
 

Stijloor

Leader
Super Moderator
Re: Gage R&R Study by Attribute

Can someone help Jean-Pierre?

Thank you very much!

Stijloor.
 

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
Re: Gage R&R Study by Attribute

AIAG 4th Ed. MSA book has a good discussion of Gage R&R for attribute. Part of the key is that the parts need to be borderline good/bad - obvious good /bad does no good. Then you analyze number mis-categorized.
 
J

JaxQC

bobdoering is right on. The parts need to be made just on the range of acceptance. Also make sure to cover all aspects ie... some gages are made to check mroe than just one criteria. You need cover for each. Good luck.
 
J

Jeenan

Sorry for the delay...I'm just back from vacation.....
Thanks for the info.......
My biggest problem, is to enter the info in the MINITAB.......Every time, it reject my entries....:mad:

Thanks!!:thanx:
 

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
We try to analyze the % of repeatability on a single operator & the % of repeatability (agreement) between the 3 operators

I would like to add that we KNOW the actual condition of the part, so it would be great to analyze if the decision (ACCEPT-REJECT-INVESTIGATE) of the operators were the good one.....

Can you provide some data so that we can help? Did you ensure there were samples that represented conditions very close to the limits, to make the test most valid?
 
Top Bottom