Comparing measurement results for a part from two ring gage

Hello All,

I think the ring gage that I am using for measuring straightness and parallelism of my part is affecting the measurement. So I measured 5 different parts using rings with different IDs. The ring with smaller ID has shown better results for part measurement. I performed Paired T-test for the comparison and found there is significance difference. I wanted to know if paired T-test can justify the ring effect on the measurement. Please advise.

Thank you
 

Semoi

Involved In Discussions
I suppose that the ring gage only yields either "fail" or "pass" as results. Is this correct? If this is the case the data is attributative/categorical and thus non-continuous. You should probably read which pre-requirements are needed for the t-test to be applied.
 
Actually the ring gage is used as a fixture and parts is placed on that.
Doubt is that using ring gage of smaller ID will yield better result.
So I measured 5 parts with two different ring and used Paired T Test to compare result and see if using ring gage of different ID can affect the measurement.

I hope now you are able to understand my question.
1641579060771.png
 

Semoi

Involved In Discussions
Do you get numerical measurement values? Do these values satisfy the requirements for the t-test to be applicable? If yes, then the only remaining question is whether or not the statistical significant result has practical relevance. As you probably know, if you have a large enough (consistent) sample size, you are able to proof statistical significance even if the difference between the two samples is tiny (=practical irrelevant).
 
Top Bottom