Hi all.

I´d like know when I can/must justify the not accomplishment of a study?

Thks for everyone.

Hello, and welcome!

If you don't mind, I think I need some help with your question.

Are you asking: " If I have performed a study and it failed, do I need to justify the failure?" or "if I performed a study and nothing was accomplished, do I need to worry about it?"

What industry are you in? I am assuming this is a gauge R&R study.

If you state that it failed, can you elaborate on how it failed?

Sorry for all the questions. If you get a chance, give us a little more information, and we'll be glad to help!

Hi, Brad. thks for u attenction!

Our auditor (ISO/TS) said me that when it will not be possible to make one study, I must justify the cause of the impossibility. If it did not have enough parts; if it would be very expensive; that is, the reason of not to have executed the study.

We assembly truck frames (chassis) and make siderails. Generality, we use to do GRR study. I have no problems with the studies I done. Perharps, I ´ll have problems with the studies [B]not done[B]. And for those cases, I must (according the auditor) tell why I not done. I am afraid haven´t a good answer that satisfy the auditor.

Thks.

Hi, Brad. thks for u attenction!

Our auditor (ISO/TS) said me that when it will not be possible to make one study, I must justify the cause of the impossibility. If it did not have enough parts; if it would be very expensive; that is, the reason of not to have executed the study.

We assembly truck frames (chassis) and make siderails. Generality, we use to do GRR study. I have no problems with the studies I done. Perharps, I ´ll have problems with the studies [B]not done[B]. And for those cases, I must (according the auditor) tell why I not done. I am afraid haven´t a good answer that satisfy the auditor.

Thks.

You used to do studies, but they were expensive. Do the results of the studies in the past degrade significatly over time when compared to one and other? If the answer is no, it seems that might be the answer. Maybe show that you do not need to do these studies because your measurement is stable due to some reason such as maintenance. Write your reasons in your procedures.

Hi, Howard!

Thanks a lot for your aid.

It´s great what you said! It´s means that if my calibration´s certificates show me that my instrument is stable I won´t need to do the study for this instrument? And can I write it in my documents as a justification? Correct me, please, if I´m wrong.

Thks again!

Hi, Howard!

Thanks a lot for your aid.

It´s great what you said! It´s means that if my calibration´s certificates show me that my instrument is stable I won´t need to do the study for this instrument? And can I write it in my documents as a justification? Correct me, please, if I´m wrong.

Thks again!

The certificate of calibration will show you the instrument is within the tolerance specified for that instrument. To determine the accuracy of your measurement system using that instrument, you will need to perform R&R study. It will be necessary to satisfy all variables in the measurement system.

Brad, you beat me to the draw.

I should have been more specific. I believe that should a succession of R&R studies prove that this element of the measurement system (R&R) is stable over time, you could probably forego more studies or extend their frequency. Maybe leave it off except for special cases such as new operators or new equipment.

I think you may be confusing the various studies. R & R is to determine the variation caused a) from repeated measurements using the equipment and b) from different operators using the equipment.

What you guys described (IMHO) are more in line with stability studies, the study of variability due to long term 'drift' of equipment caused by use etc. Dependingon the equipment type, linearity and bias studies are often performed too, to characterize the measurement system.

I agree with Andy - calibration data over time only shows stability of the equipment. It doesn't address all of the variables in the measurement system. Using stability data only would completely ignore variability due to operators.

I agree with Andy - calibration data over time only shows stability of the equipment. It doesn't address all of the variables in the measurement system. Using stability data only would completely ignore variability due to operators.

I'm just curious, barring a change in operators and training, what would change the variability due to operators? Once the variability of your operators has been established, how frequently would you repeat an R&R study? If the study is conducted using the same operators over and over, as the OP suggests, would the operator variation change significantly?

Hi, Andy.
Thanks for your help too.

Dear Mnavarro:

The Gage R&R is perform to get an estimate of the precision of your measurement system.

In other words, if your measurement system is perfect and you measure the same part multiple times, you will always get the same number. In practice, since measurement systems are not perfect, the observed variation coming from measuring the same part will provide you with an estimate of the measurement system error (During a real gage study you use several parts in order to improve the confidence interval of your estimate).

When you calibrate a measurement system, you are adjusting its bias. In other words, you are adjusting its accuracy.

General Note:
When you don’t have a gage R&R study, you don’t have any information with regards the precision of the measurement system. The implications are uncertainty in your pass/fail criteria. You could end up accepting bad parts and rejecting good parts.

The only way that you could justify not doing a Gage study is to somehow show objective evidence that you are protecting the customer.

I think you may be confusing the various studies. R & R is to determine the variation caused a) from repeated measurements using the equipment and b) from different operators using the equipment.

What you guys described (IMHO) are more in line with stability studies, the study of variability due to long term 'drift' of equipment caused by use etc. Dependingon the equipment type, linearity and bias studies are often performed too, to characterize the measurement system.

I think you may have clarified a concern of mine - but I'd like to confirm. I have a couple of measurement processes on my control plan that don't fit into a nice R&R, and I've been looking for an adequate explanation to address this.

One is a fixed ircon measuring strip temperature at a critical point on our galv line. We have evidence of acceptable linearity and bias over multiple calibrations. These calibrations also indicate there are no concerns with stability. So... if I have demonstrated via the calibration / stability protocol that there is no issue from "repeated measurements over time" and by nature of the device (permanently fixed gauge) there is no effect from "different operators using the equipment" -- can I not stop there in the MSA process?