# Reproducibility – exam question

Six Sigma – iSixSigma Forums Old Forums General Reproducibility – exam question

Viewing 18 posts - 1 through 18 (of 18 total)
• Author
Posts
• #45188

I’Anson
Participant

Hello all,
I have taken an exam and I’m not comfortable with the answer given.
Here is the question:
Data is being collected over the course of a day and during two different shifts.  Operators in the second shift are using a different measurement device than the first.  What type of measurement error is this?
a)Calibration,  b)Repeatability,  c)Reproducibility, d)Other

0
#146722

Deep
Participant

Daniel:d) Other,,, i guessDeep

0
#146723

Mikel
Member

It is a very poorly worded question, but the answer will be c).
The reason it is poorly worded is that the difference using two different devices, if a difference exists, will show up as reproducibility.

0
#146725

Participant

Must be an ASQ exam question.

0
#146727

Monk
Participant

Daniel,
This has to be a)  calibration error.
The reason being that the resulting measurement error will be due to variation in the calibration of the two measurment devices being used.
Monk

0
#146728

Mikel
Member

Wrong

0
#146729

Monk
Participant

Reason ?

0
#146730

Tony Bo
Member

Stan is right, the question is..what kind of measurement error?………the measurement error type is reproducibility…it will show up as reproducibility.

0
#146731

BTDT
Participant

Daniel:Yes – a poorly worded question, but the answer would probably be (c) reproducibility.If we were to do an ANOVA on the data collected, we would get some number for total variation (SS,total). In order to see the effect of operators (reproducibility) I would compare the results from the first shift against the second shift (SS,between). The error owing to the equipment should be what is left over (SS,total – SS,between) = SS,within, but this assumes the same piece of equipment is used by both shifts.The problem is that the ANOVA doesn’t know that both the operator AND the equipment has changed when the calculation for SS,between was done. Any error owing to differences in equipment will end up looking like operator error.In the bigger picture, the two shifts are producing different numbers because there is variation between their procedures. It just happens that equipment changes are included in their different procedures.Does that make sense?Cheers, BTDTBTW – what answer is the correct one according to the examiners?

0
#146733

Torrance
Participant

Hi Monk – I agree with Stan. (God help me..  :-))
Daniel, from the title of this thread, looks to me like you got this one right. (But unfortunately for you, neither Stan or myself is marking your exam paper, and despite our opinions, you’ll probably find that the marker wont consider our thoughts…. lol)
Monk: Fully understand your point about calibration, so to explain my own thoughts.
We can say reproducibility from the facts given. I.e. we are using different equipment.
Its difficult to take things as far as stating calibration error, as the equipment could be calibrated perfectly, while we still get variation. (e.g. different operators, something else) We dont want to jump to conclusions without further analysis (which we cant do in an exam).
Stupid question – not really sure why the answer is important. In a real life situation – we would look at obvious differences, and fix them first. Seems simple to me – use same equipment on each shift – still get the variation??….Yes – lets discuss further. No – Monk, you could be right after all!!
The guy who set this question, will probably never get past the debate on what type of error it is – just one more reason why lots of companies lose faith in Six Sigma….
To be honest, it could be reproducibility, calibration or Other – could be time for one of the most important and effective tools – JFDI. Just F’#ing Do It.

0
#146734

Mikel
Member

It’s reproducibility.
Even if there is a calibration problem, it will show up as a difference in the two measurement systems (reproducibility – by definition). The solution to the issue, if one actually exists, may be to address calibration, it might be training, it might be fixturing, …..
And this is not an indictment of Six Sigma. MSA is a tool in the Six Sigma toolset, but is not exclusive to Six Sigma. It is a poorly worded question and nothing more.

0
#146739

Monk
Participant

Stan,
As per MSA definition, for ‘reproducibility’ error, both the operators need to use the same measurement device. In this case it is not the same…..so how do we call it a reproducibility error….it has to be ‘calibration’ error or ‘others’.
Monk.

0
#146741

Mikel
Member

That definition is incorrect. We are trying to guarantee that all iterations of the same measure agree. It has nothing to do with the same device or not. One of the best MSA study that we can run is agreement between our measurement and our supplier’s or customer’s measurement of the same thing. There certainly are different devices in use.

0
#146744

BB06
Participant

The answer will be “C”. Its always reproducibility when you are measuring “under different conditions”
Hope this helps.

0
#146745

Taylor
Member

I agree with Davey T.  What does it matter if you know the proper name to call the problem.  In the real world you fix it.

0
#146746

Cravens
Participant

If you take Reigle’s advice and use “a rule of thumb”, then the error is caused by different lengths of thumbs.

0
#146747

LeDrew
Participant

I think the ansswer would be reproducibility but the cause of this reproducibility would be a difference of calibration between the two instrument…
but that still a question of a poor teacher.

0
#146748

Mikel
Member

Calibration is only one of many possible causes

0
Viewing 18 posts - 1 through 18 (of 18 total)

The forum ‘General’ is closed to new topics and replies.