Define the term Sigma in Six Sigma
Six Sigma – iSixSigma › Forums › Old Forums › General › Define the term Sigma in Six Sigma
 This topic has 11 replies, 8 voices, and was last updated 15 years, 8 months ago by BritW.

AuthorPosts

December 4, 2006 at 9:55 am #45415
I would like to know what is the significance of the term “Sigma” in Six Sigma
0December 4, 2006 at 11:24 am #148376Naveen…I had a so called smart butt black belt once ask me to define Sigma…and i asked which one?Normally Sigma in statistics is the Standard Deviation of the Population or basically the spread of the values about an average or middle value (mean, mode, medium or geometric). Therefore you tend to have a + and a – value. e.g. +/ 3 Sigma (but you may have +4 and 2 for skewed distributions)…which is 6 in total and where Six Sigma(TM) comes from.In Six Sigma(TM) talk, it is a similar thing, but you start at 0 and everything is plus (+). Like the cumulative of the values so you remove the negative s and a comparison between Sigma’s can be made (In theory)That said i have work at some so called leading companies where BB’s and MBB’s get negative Six Sigma(TM) Sigma values, which is impossible. The reason…because they use an Excel equation called INVNORMDIST! Which is wrong.Boring stuff, but if you’re feeling suicidal one day and need a distraction, you can plot this Excel equation and compare it to the Yield values that Six Sigma(TM) uses. You will see that they do not correlate at all, except at 50% defect rate.That said, today the sun is shining bright and i would not waste your time. Enjoy the day. Life’s too short.
0December 4, 2006 at 12:10 pm #148378“but you start at 0 and everything is plus (+). “
What in the hell are you talking about ?
Total crap !!!0December 4, 2006 at 12:49 pm #148379Really.
That is as bad an answer as I have ever seen on here.
Naveen – read the new to Six Sigma? in the blue bar to the left.0December 4, 2006 at 2:13 pm #148383Oh my. Not sure of your backgraound Alex, but please read the New to Six Sigma link on the left and search for sigma level, DPMO, et al. I hope you will explain it a bit differently next time you are asked. Your response was wrong in that sigma level has nothing to do with the +/ 3 sigma control limits as you described in your first full paragraph.
If you are a black belt – get your money back.
0December 4, 2006 at 7:40 pm #148405That is just as bad a reply as that previous idiot.
There is only one meaning of sigma … the one used for control limits.0December 4, 2006 at 7:56 pm #148411Steve:
Sorry for your confusion and inability to read into my response. There is sigma (standard deviaiton) and sigma level:
A Sigma Table (also called a Z Table) is used to determine the number of standard deviations from the mean necessary to equal the DPMO defect rate (i.e., the sigma level). The smaller the defect rate, the larger the Sigma level or Z score.
Hope this helps0December 4, 2006 at 8:05 pm #148413I suppose what I should have said explicitly was that his assertion that the “6” in 6 Sigma came from the +/ 3 sigma control limits he described below was wrong. 6 refers to a measure called sigma level which refers to the Ztable. This is a count of the number of standard deviations that can fit between a mean and a target value. +/ 3 Control limits refer to the Shewart Control Chart limits placing a UCL and LCL at 3 standard deviations from the mean. Sigma Level and Sigma are not the same thing.
Normally Sigma in statistics is the Standard Deviation of the Population or basically the spread of the values about an average or middle value (mean, mode, medium or geometric). Therefore you tend to have a + and a – value. e.g. +/ 3 Sigma (but you may have +4 and 2 for skewed distributions)…which is 6 in total and where Six Sigma(TM) comes from.
Please excuse my poorly worded response from the last post.0December 4, 2006 at 9:02 pm #148419Bright eyes,
Whats the difference between sigma in a six sigma table and sigma in control charts ?0December 4, 2006 at 9:20 pm #148422Hal – I know you know the difference between sigma level (from the standard Z tables) and standard deviation (sigma).
Not trying to pick are you?0December 4, 2006 at 10:16 pm #148430Britw,
Hal is just making fun of you. Everybody here knows that the 1.5 shiftdriftcorrection is rubbish and six sigma tables should never be used. Values of sigma from these tables are total nonsense.0December 5, 2006 at 2:15 pm #148451Thanks Phil. I certainly wasn’t trying to get into that argument. Just wanted the previous poster to know that there was a difference.
0 
AuthorPosts
The forum ‘General’ is closed to new topics and replies.