Variance
Six Sigma – iSixSigma › Forums › General Forums › Training › Variance
 This topic has 15 replies, 7 voices, and was last updated 11 years, 9 months ago by MBBinWI.

AuthorPosts

December 9, 2010 at 5:56 pm #53668
ChauffeParticipant@mackcc59 Include @mackcc59 in your post and this person will
be notified via email.In the event I have a standard deviation of 0 (zero), no variance, what should I ask myself and what will the Pp, Ppk values be?
0December 9, 2010 at 6:28 pm #191028Get a better measurement system. There is no such thing as 0 variance.
0December 9, 2010 at 6:32 pm #191029I agree. If you are seeing no variance whatsoever then either your process is flawless or your measurement system needs some attention as it is unable to detect the variation.
0December 9, 2010 at 6:39 pm #191030
ChauffeParticipant@mackcc59 Include @mackcc59 in your post and this person will
be notified via email.So if my Upper Spec is 140 and Lower Spec is 100, and all data measurements are 115, SD is 0, are the Pp, Ppk values NA?
0December 9, 2010 at 6:41 pm #191031I think what Stan and I meant are that the computing the Pp and Ppk would be senseless because the measurement system is not adequate and should not be trusted.
0December 9, 2010 at 6:54 pm #191032
ChauffeParticipant@mackcc59 Include @mackcc59 in your post and this person will
be notified via email.Thanks Guy’s!!!
0December 10, 2010 at 8:31 pm #191033Stan, you tard.
Why are you sendingMack of on a measurement boondoggle. Of COURSE there’s no such thing as zeroo variance, but there is such a thing as an MSA %study variance and a Gage r&R. His measurement system may be just fine.
Mackie, use 6 for your Ppk and dare your customer to challenge you.
Peace.
0December 11, 2010 at 12:43 pm #191034Miker,
It’s retard, not tard.
That’s like calling you a fus instead of the dufus you really are.
You can’t pass GR&R with 0 variance in your sample and if his current system says there is no variance that has to be what his sample is.
The advice to claim 6 is just stupid. It’s always great to see advice coming from people who don’t know Jack.
He needs to have discrimination in his measurement system, he has none.
0December 13, 2010 at 3:46 pm #191035
SchwarzParticipant@DaveS002 Include @DaveS002 in your post and this person will
be notified via email.There is such a thing as zero variance.
An hour ago I counted the money in my wallet. It was $21. Thirty minutes ago, I counted it again. Still $21. Just now, I counted it again. Still $21.
Over that hour period, the variance of the amount of money in my wallet is zero.
Of course, the variable of interest here is not stochastic and so it doesn’t fall under the general definition of things that are studied and discussed in this forum. My point is that ‘nonchanging’ entities, measured without error, can have zero variance.
0December 15, 2010 at 11:26 am #191036Dave, you are right. What a deep and heady answer.
It’s like if I measured the value of your answer 20 times. It had no value all 20 times. Zero variance!
Now to the useful world and process improvement. There is no such thing as zero variance.
0December 15, 2010 at 2:52 pm #191037
SchwarzParticipant@DaveS002 Include @DaveS002 in your post and this person will
be notified via email.Stan, Of course you understand that my reply was an intellectual exercise and not advise to the forum or the original poster. It points out that some processes can be devised that do not vary over the measurement period. Whether to call the variance zero or undefined can be debated.
Obviously in quality measurement scenarios involving stochastic processes, the runtorun variance is not zero. Sample data that is identical for each measured value is probably due to a badly designed test or defective measuring instrument.
Once a year I survey the maturity level of this forum and community. I see it has not improved. See you next year.
0December 16, 2010 at 1:27 am #191039
MBBinWIParticipant@MBBinWI Include @MBBinWI in your post and this person will
be notified via email.DaveS002 wrote:
There is such a thing as zero variance.
An hour ago I counted the money in my wallet. It was $21. Thirty minutes ago, I counted it again. Still $21. Just now, I counted it again. Still $21.
Over that hour period, the variance of the amount of money in my wallet is zero.
Of course, the variable of interest here is not stochastic and so it doesn’t fall under the general definition of things that are studied and discussed in this forum. My point is that ‘nonchanging’ entities, measured without error, can have zero variance.
I’m with Stan on this one – your answer was just plain stupid.
Everyone knows you need three operators to do a GR&R :laugh:
0December 16, 2010 at 1:31 am #191040
MBBinWIParticipant@MBBinWI Include @MBBinWI in your post and this person will
be notified via email.Methinks this was a homework question. You should do your own work.
If you want to submit a reasoned out answer and ask if you’ve followed a logical/appropriate path, that’s one thing. But don’t just sling out your homework question and expect us to answer it for you.
Jeesh. You’d think that people would get a clue.
0December 16, 2010 at 1:32 am #191041
MBBinWIParticipant@MBBinWI Include @MBBinWI in your post and this person will
be notified via email.On second thought, let me measure your money. I bet that it will be a different value!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
0December 27, 2010 at 8:41 pm #191084To answer the original question, Pp and Ppk would typically be defined as n/a since they can not be calculated when the standard deviation of the individual values = 0. (You can not use 0 as the denominator and get a number. Your calculator would show this as an error and Excel would list the answer as “#DIV/0!)
0December 28, 2010 at 3:23 am #191085
MBBinWIParticipant@MBBinWI Include @MBBinWI in your post and this person will
be notified via email.Matt wrote:
To answer the original question, Pp and Ppk would typically be defined as n/a since they can not be calculated when the standard deviation of the individual values = 0. (You can not use 0 as the denominator and get a number. Your calculator would show this as an error and Excel would list the answer as “#DIV/0!)
Actually, the answer, using your logic, would be infinity, since anything divided by zero is infinite.
This was just a sloth trying to get others to do his thinking for him (or his homework, which amounts to the same thing).
0 
AuthorPosts
You must be logged in to reply to this topic.