MSA for Attribute Data System - We have a gauge called ?Focus?

D

darkafar

MSA for Attribute Data System - We have a gauge called “Focus”

We have a gauge called “Focus”. When we use Focus to measure the object, Focus gave an image. By comparing the image with standards, the operators gave a grade, numbering 1, 2, 3, and 4.

Our customer thinks the parameter is important therefore it is listed on the control plan. According to TS16949 we should do a MSA to this gauge. My problem is how to do this MSA?
 

antoine.dias

Quite Involved in Discussions
Re: MSA for Attribute Data System - We have a gauge called “Focus”

IMO it is perfectly possible to do attribute MSA as described in the manual.

-) You have to check if the same operator is giving the same grade for the
same part if the part is checked 3 times ( repeatability )
-) You have to check how other operators do the same ( reproducibility )
-) You have to check how the 3 operators are giving the correct grades in
relation to the reference or what the customer wants ( effectiveness,
false alarm, missed parts )

Best regards,

Antoine
 
A

Atul Khandekar

Re: MSA for Attribute Data System - We have a gauge called “Focus”

You may also want to look up:

1. Kendall's Coefficient of Concordance.
2. Weighted kappa method


Disclaimer: I have not really studied and never used these methods. I hope some of the forum statistics experts would chip in with their valuable opinions on which (if any) of the above methods can be used in this case.
 
Top Bottom