Intro to MSA of Continuous Data – Part 7: R&R using Wheeler’s Honest Gage Study

[antoine.dias]

Hi,

I 've just found part 1 to 7 on MSA.

Wow, What a good piece of work.

Thanks a lot for the information.

Antoine Dias

 

[Guest]

This may have an obvious answer I am not seeing, but I am wondering if the calculation of the probable error on the ANOVA worksheet is accurate. Wheeler writes this to be 0.675*R&R. In the worksheet, in each of the ANOVA tables it is simply taken to be 0.675*EV. For ANOVA tables 1 and 2, do we not also need to consider the AV?

 

[Miner]

This may have an obvious answer I am not seeing, but I am wondering if the calculation of the probable error on the ANOVA worksheet is accurate. Wheeler writes this to be 0.675*R&R. In the worksheet, in each of the ANOVA tables it is simply taken to be 0.675*EV. For ANOVA tables 1 and 2, do we not also need to consider the AV?

Please refer to Wheeler's article entitled Relative Probable Error. Wheeler defines P.E. = 0.675 σ^e. σ^e is the variation due to repeatability. Remember, this is the probable error of a single measurement by a single operator. I also confirmed all results using Wheeler's data set to Wheeler's results.

 

[Guest]

Please refer to Wheeler's article entitled Relative Probable Error. Wheeler defines P.E. = 0.675 σ^e. σ^e is the variation due to repeatability. Remember, this is the probable error of a single measurement by a single operator. I also confirmed all results using Wheeler's data set to Wheeler's results.

I think the nomenclature of σ^e is in direct contradiction to the use of the same term in "An Honest Gauge R&R Study" (see Steps 2 and 4). I will admit that he does find σ^e in his examples using a method that would only identify repeatibility. However, he uses σ^pe in his other paper to identify repeatability. Being that PE is the probable error of a single measurement, I can see how this would essentially be the repeatability of the system, and so I do agree that the math is accurate.

 

[Guest]

Thanks!, your excel file is excellent.

 

[refractoryman]

So how does one determine if there is interaction or no interaction when wanting to select the ANOVA table? Also, I could use help understanding what the Alpha values to remove interaction and pool between operator mean. You are dealing with someone who knows nothing about all of this and is desperately trying to learn on his own. Thank you in advance for your kind help.

 

[Mr.Happy]

Hi Miner,

I saw that you made an Updated 7th March 2014
Is the attached Wheeler's Honest Gauge Study.xls also updated to the latest AIAG MSA manual (4 th. Edition)?
Also I have a question about P/T Ratio (one of our customer’s expectation) I mis this in the attached Wheeler's Honest Gauge Study.xls? Could you give us some more information about this P/T Ratio?

 

[Marc]

Links in the starting Blog Post have been verified and work.