Help to calculate sigma value
Six Sigma – iSixSigma › Forums › Old Forums › General › Help to calculate sigma value
 This topic has 8 replies, 6 voices, and was last updated 17 years, 8 months ago by Ron.

AuthorPosts

January 22, 2004 at 10:07 am #34351
Shalini MudaliyarMember@ShaliniMudaliyar Include @ShaliniMudaliyar in your post and this person will
be notified via email.Hi All,
Am new to this site and found the same very informative.
I am new in this field and would like to have some help in calculating sigma value for a critical process.
I have 11 units measured across 5 satisfaction levels (1 being the best and 5 the least).
Please let know which is the best way of doing the same.
Regards
Shalini
0January 22, 2004 at 1:18 pm #94485Shalini,
Usually we first map our processes (high level and low level) and calculate the defect per each step. Then calculate the RTY (which is the multiplication of defects/total # produced per each step). Having the RTY you can check in the Ztable what is the corresponding area under the normal curve and you can get your sigma value. To get more exact calculation you may use Excel with the function NORMSINV
If you need some more detailed information please let me know
hope this helps
Anna0January 22, 2004 at 1:34 pm #94486
indreshParticipant@indresh Include @indresh in your post and this person will
be notified via email.hi!
just apply the following logic
– see if we give weightage to 15 qualitative parameters from 2, 1,0,1,2 respectively then the maximum score that one can have is 22 across 11 different variables and minimum 22
now at 22 the defect is zero that is the desired level and below and at 22 its totally defective
secondly the 11 variables criticality can also be one clear criteria to have defects counted on…..you may take the Voc( voice of the customer that below a certain criterion he will not accept the output) that will be the defect level
check out the applicability and see if it suffice
rgds,
indresh
i can be contacted at [email protected]0January 23, 2004 at 4:02 am #94543
indreshParticipant@indresh Include @indresh in your post and this person will
be notified via email.copy this on excel
column A to K
Write under column A to K following
– Process
– defect
– unit
– opprotunity
– total opportunity formula =IF(C3, C3*D3, “”)
– DPU formula =IF(C3, B3/C3, “”)
– DPO formula =IF(C3, B3/E3, “”)
– DPMO formula =IF(C3, G3*1000000, “”)
– Shift under shift write constant 1.5
– Long term capability formula =IF(E3=””,””,IF(B3=0, NORMSINV(EXP(LN(0.5)/E3)), SQRT(LN(1/(G3)^2))((2.515517+0.802853*(SQRT(LN(1/(G3)^2)))+0.010328*(SQRT(LN(1/(G3)^2)))^2))/((1+1.432788*(SQRT(LN(1/(G3)^2)))+0.189269*(SQRT(LN(1/(G3)^2)))^2+0.001308*(SQRT(LN(1/(G3)^2)))^3))))
see if you generate the excel sheet
cannot send you the attachement due to internet security issue
see if it suffice else revert back to me
rgds,0January 23, 2004 at 4:05 am #94544
indreshParticipant@indresh Include @indresh in your post and this person will
be notified via email.the last column K will be a sum of the short term capability value and the column I where you have written constant as 1.5
rgds,indresh0January 23, 2004 at 6:00 am #94547Hi Indresh,
Thanx a lot for your response .
Would require some more clarification from you. I wish to calculate the sigma value for a VOC process , for which I have 11 parameters. There might be cases where all the 11 questions are not answered by the customers .Hence for a single form , of the 11 responses reqd , I can get reply for all 11 or anything less than that.
I have a total of 190 responses.Hence my total unit becomes 2090.
So now can you suggest what should be done.
Anna,
Would like to thank you as well for the response. But can you kindly explain the concept of RTY.
Regards
Shalini0January 23, 2004 at 6:34 am #94548
indreshParticipant@indresh Include @indresh in your post and this person will
be notified via email.hi,
its just that the opportunities are getting reduced
let me clearify with an example
– if you have given the questionnaire with 11 parameters to a customer
he has given response of only 8 questions
so here the unit (that is one questionnaire =1) and opportunities would be 8 and not eleven
so for total number of units total opportunities might not be unit*11 but lower than that
the sigma would thus automatically increase
rgds,
indresh
0January 23, 2004 at 9:13 am #94549
Manikanda Mariappan SParticipant@ManikandaMariappanS Include @ManikandaMariappanS in your post and this person will
be notified via email.Let me clarify what you are looking for….You are trying to quantify the effectiveness of the VOC collection method using a Z score, right ?
If I’m right, then follow this….
What you need is a different form of data from the 190 response sheets 11 questions.
1. Identify the number of questions answered by the customers in each response sheet, build them as a frequency table and summarize it. That is, write down how many response sheets had only 1 answer, only 2 answers and so on till 11 answers.
The table should look like this for eg…
only 0 ans5, 1 ans10, only 2 ans25, only 3 ans40, only 4 ans60 and so on.
The total of all the values should be 190 as you have only 190 responses (as per your post) [These values are only the questions answered, for question not answered (defects) you need to use 190 minus this data.] So you have a set of attribute data (poisson distribution).
2. Now set an acceptance limit for the number of answers. For eg, if you think 8 or more answers is an acceptable, then total all values from 8 ans to 11 ans. 190 minus this value will be your defects. Here defect means the number of responses with answered questions below the acceptance level chosen, in this case it is 8. Now you got the defects, your unit size is 190, oppurtunity is 1 per unit. Using this calculate the DPMO (defects per million oppurtunities) and from the Z table or Excel you can get the Z score.
3. You can set a different acceptance level and calculate the Z score for different acceptance levels. The Z score will tell you the effectiveness of your VOC procedure with your customers.
Thanks,
Manikanda Mariappan S0January 23, 2004 at 12:37 pm #94555A sigma calculation is easily implemented:
1st: Define what a defect is, remember to differentiate between a defect and defective.
2nd. Calculate DPMO
3rd Convert DPMO to Sigma there are many tables out there you can use if you can’t find one post again and I can send you one.
Never calculate sigms from an RTY value; that pratice will disguises the true nature of each step in the process. RTY is utiized to highlight an entire process to indicate its need for improvement.
Defects are normally operationally specific or at final inspection (typically the later).
RTY is utilized to uncover the hidden fatory of rework and scrap not normally observed at the final inspection point.
0 
AuthorPosts
The forum ‘General’ is closed to new topics and replies.