cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Choose Language Hide Translation Bar
BurnStats
Level I

Percent agreement for rater calculation in attribute gauge study

Hello.

Can anybody explain what is meant by "number of uncounted matching levels for this rater k within part i for rep j" in the "Statistical Details for the Gauge Attribute Chart" section of the JMP Documentation Library (page 4760 out of 7278)? Also how the example calculation using the data in Table 7.1 is done? Screenshots of the relevant JMP Documentation Library pages are attached.

Thank you very much.

3 REPLIES 3
jthi
Super User

Re: Percent agreement for rater calculation in attribute gauge study

This is how I understand "number of uncounted matching levels for this rater k within part i for rep j" based on the example calculations ( Quality and Process Methods > Attribute Gauge Charts > Statistical Details for Attribute Gauge Char... ) but I could be totally wrong and just got lucky.

F or A and B rater

jthi_0-1698601778838.png

How many ones there are

 

jthi_2-1698601838484.png

Still how many ones are, but now we have already counted one 1 in step 1

 

jthi_4-1698601871725.png

How many zeros there are

 

And similarly for C:

jthi_5-1698601970655.png

jthi_6-1698601977361.png

jthi_7-1698601980190.png

 

-Jarmo
BurnStats
Level I

Re: Percent agreement for rater calculation in attribute gauge study

Hi. Thank you. This is very helpful, although I do not understand the logic of the process. If JMP could provide a reference, that would be useful.


Regards.

statman
Super User

Re: Percent agreement for rater calculation in attribute gauge study

The JMP help is not meant to teach statistical methods (or the reasoning behind the different methodologies).  That being said, if you want to understand the methodology search Inter-class correlation and Kappa.

 

This is the JMP help section:

https://www.jmp.com/support/help/en/17.2/#page/jmp/statistical-details-for-the-agreement-report.shtm...

 

"All models are wrong, some are useful" G.E.P. Box