Intro to MSA of Continuous Data – Part 6: R&R for Non-Replicable Measurements

Rameshwar25

Quite Involved in Discussions
#13
Miner, Thank you for sending me the link. I have, too some extent, understood it and have gone through some examples from minitab. I am ready to prepare a excel sheet but few questions are still there.
1. There is no two different tables (one for interaction and one for without interaction). There is only one table involved. Is it correct?
2. F Values is not compared with P-value to see if there is interaction. Is is correct?
3. If answer to quest 2 is correct then what is the use of calculating F value in nested anova method.


Regargs
Rameshwar
 

Miner

Forum Moderator
Staff member
Admin
#14
Rameshwar25;bt726 said:
Miner, Thank you for sending me the link. I have, too some extent, understood it and have gone through some examples from minitab. I am ready to prepare a excel sheet but few questions are still there.
1. There is no two different tables (one for interaction and one for without interaction). There is only one table involved. Is it correct?
2. F Values is not compared with P-value to see if there is interaction. Is is correct?
3. If answer to quest 2 is correct then what is the use of calculating F value in nested anova method.


Regargs
Rameshwar
Sorry for the delay, but I do not check comments on my blog on a regular basis.

1. That is correct. When you have a nested setup, interactions cannot physically exist between the different levels of nesting.
2. Again, the Operator*Part interaction physically cannot exist, so it cannot be tested.
3. It is still of value to determine whether the Part or the Operator terms are statistically significant.
 

Rameshwar25

Quite Involved in Discussions
#15
Miner, Thanks for the reply.
I studied some examples from Minitab and formula given in the pdf file for which you sent me the link. I got some understanding, prepared an excel file for this purpose and posted it on 15th May 11. The link is given here.
GRR (Gage R&R) Nested Anova Method Example
Although, the results match with Minitab, I request you to critically check the contents and to suggest to improve the same.

Regards
Rameshwar
 
I

isotsadmin

#16
does anybody have an excel spreadsheet for non replicable msa one sided tolerance? appreciate your help
 

Marc

Captain Nice
Staff member
Admin
#17
"does anybody have an excel spreadsheet for non replicable msa one sided tolerance? appreciate your help"

FYI - You will probably get more "eyes" on your request if you also ask in the forum.
 

AYDOSE

Starting to get Involved
#18
Miner, first of all I would like thank you for this clear and detailed identification.
We made MSA study, Gage R&R nested ANOVA for elongation using minitab. Our results are as folllows:
Study Var %Study Var
Source StdDev (SD) (6 * SD) (%SV)
Total Gage R&R 1,00399 6,0240 33,22
Repeatability 1,00399 6,0240 33,22
Reproducibility 0,00000 0,0000 0,00
Part-To-Part 2,85091 17,1054 94,32
Total Variation 3,02253 18,1352 100,00
 
Number of Distinct Categories = 4

I know from Gage R&R NDC shold be greater than 5 but in our calculation is 4 .
If I can take your comments about the resuls I will very much appreciated. Thanks, for your help
 

Miner

Forum Moderator
Staff member
Admin
#19
Sorry for the long delay in responding. Vbulletin did not alert me to comments on my blogs, so I did not see this until today. Minitab Help has an explanation of the formulae used in a nest R&R study.
 

Top Bottom