Thanks for the very informative guidelines but only one query for six sigma limit , is it not the formula (μ±3σ) ???? , instead in this video it is taken as (μ±6σ) please suggest.
Good day Stats, I'm impressed on this tutorial that much , that I just going to follow the exercises once again for better understanding, but I have some questions, there is some duplicates on the data, and also I wonder if I need to sort the data from the largest to smallest. Please advise. Thank you so much for creating this video
What's means "bins" I'm not sure.. if i have tolerance for example from 102,2 to 107,2 and my six sigma limits are 96,13 and 113,09, i have to include six sigma limits, so my bins should be from 96,13 to 113,09, and i have thirty measures it is my "data", and what's bins i should take to my calculation ? For example 96,00 - 96,40 - 96,8 etc ? Or 96 - 97 - 98 ? And why, thank you for answer
Here the video has a mistake... instead of calculating for 6 sigma the video calculates for 12 sigma its (mue) + 3 (sigma) and (mue) - 3 (sigma)... 3 sigma one side of mean and 3 on the other side makes a six sigma
I am a CSSGB and CQPA. You are literally categorically wrong. Six sigma means 6 deviations between the average and BOTH the USl and LSL. This is why you end up with 3.4 ppm and not ~0.27%. You are referring to a normal plot.
For six sigma limits, use the following formula instead as it was difficult to get a perfect match for the bin value. Also, limit the number of bins you create as the more refined it is in decimal places the more number of six sigma matched results it will give. I did this by limiting to 50 bin arrays. =IF(ROUND(E68,2)=ROUND($F$63,2),$F$65,IF(ROUND(E68,2)=ROUND($F$64,2),$F$65,0))
Hi, obtaining the 6 sigma SL's become very difficult as the precision of measurement increases. Say when you are working in inches, and your measurements range in 3 decimal digits, say .235", the bins would have more than a single cell which can round off close to .235" value. This will result in multiple SL's line in the chart. Also, it is best to take min of LSL or LSL of 6sigma and max of USL or USL of 6 sigma to arrive at the bins value. If the mean offset from SL's, then you cant have the tolerance limits in the chart. Let us know you have a solution for the above.
I don't know what I am doing wrong but I am trying to calculate 3 sigma and it just is not working using this excel video, I am getting all zero's for 3 sigma
This tool does not work for my situation, I have product failure related projects, but there are no data. I also learned that the statistics works with data, if no pertinent data statistics would not work. If you are trying to use this technique in a manufacturing environment, this may not yield anything, for sure it is your money you have the right spend the way you wanted it. There are better ways to handle problems like this.
Mr. Uday Nair, Yes, you are very correct, attributes of failures are many, like some of them mentioned above and it would be very difficult to pick one of the attributes and collect data from the manufactured parts and detect the cause of failure, after doing all those exercise there is no defect found, and you may realize to look for some other characteristics and keep doing the exercise until you find a margin of error in one of the attributes. By the time you do the exercise to make the Six sigma applicable, few things are going to happen 1. You spend lot of time in collecting the data and analysis until a defect is detected, 2. Cost, cost of inspection process or Six Sigma process would exceed the cost of the product itself 3. The defect identified may not be a critical one or may be minor one and it would be acceptable as is and 4. What if the sample size or production lot is very few, not enough samples to draw data, what would be the next action? Not to mention that you would pass the delivery date of the product, your customers are waiting to get the parts to proceed with the next operation, the bottom line is the risk is much more than the benefits. I went through these exercises earlier and from the experience I am speaking all these. Thank you for bringing your suggestions for discussion sake and hope this would shed some light in the subject matter we are in discussion and it may help some readers of this post. You or any one can contact me to know more about the process control, do not get me wrong this Six Sigma process is a good tool or technique in a product or process where there are no basic or datum or for some developmental projects to define the product or process. Contact: bmatthewbob@gmail.com
@Uday Nair Mr. Uday Nair, Yes, you are very correct, attributes of failures are many, like some of them mentioned above and it would be very difficult to pick one of the attributes and collect data from the manufactured parts and detect the cause of failure, after doing all those exercise there is no defect found, and you may realize to look for some other characteristics and keep doing the exercise until you find a margin of error in one of the attributes. By the time you do the exercise to make the Six sigma applicable, few things are going to happen 1. You spend lot of time in collecting the data and analysis until a defect is detected, 2. Cost, cost of inspection process or Six Sigma process would exceed the cost of the product itself 3. The defect identified may not be a critical one or may be minor one and it would be acceptable as is and 4. What if the sample size or production lot is very few, not enough samples to draw data, what would be the next action? Not to mention that you would pass the delivery date of the product, your customers are waiting to get the parts to proceed with the next operation, the bottom line is the risk is much more than the benefits. I went through these exercises earlier and from the experience I am speaking all these. This is from your note “of course not saying this will be conclusive but can be helpful”, how much helpful? Spending the amount of time and effort would justify the cost of your finding, a very minor one; one of my instructors told me that Statistics is value neutral, no product or process knowledge. Thank you for bringing your suggestions for discussion sake and hope this would shed some light in the subject matter we are in discussion and it may help some readers of this post. You or any one can contact me to know more about the process control. Do not get me wrong this Six Sigma process is a good tool or technique in a product or process where there are no basics or datum or for some developmental projects to define the product or process. Contact: bmatthewbob@gmail.com. Regards.
Thank you very much. It took me 2 days to find a way to do it researching and finally i found everything i needed in one place :).
Well explain ,Thanks for the video.
Thanks for the very informative guidelines
but only one query for six sigma limit , is it not the formula (μ±3σ) ???? , instead in this video it is taken as (μ±6σ) please suggest.
it should be (μ±3σ), here the video calculates for 12 sigma which can never be achieved
Thanks for the very informative guidelines.It is very help us.I am thankful to you.
Good day Stats, I'm impressed on this tutorial that much , that I just going to follow the exercises once again for better understanding, but I have some questions, there is some duplicates on the data, and also I wonder if I need to sort the data from the largest to smallest. Please advise. Thank you so much for creating this video
Best process capability explanation for excel. Thank you...
What's means "bins" I'm not sure.. if i have tolerance for example from 102,2 to 107,2 and my six sigma limits are 96,13 and 113,09, i have to include six sigma limits, so my bins should be from 96,13 to 113,09, and i have thirty measures it is my "data", and what's bins i should take to my calculation ? For example 96,00 - 96,40 - 96,8 etc ? Or 96 - 97 - 98 ? And why, thank you for answer
Here the video has a mistake... instead of calculating for 6 sigma the video calculates for 12 sigma its (mue) + 3 (sigma) and (mue) - 3 (sigma)... 3 sigma one side of mean and 3 on the other side makes a six sigma
I am a CSSGB and CQPA. You are literally categorically wrong. Six sigma means 6 deviations between the average and BOTH the USl and LSL. This is why you end up with 3.4 ppm and not ~0.27%. You are referring to a normal plot.
Many thanks from 0 to full excel spreadsheet about cpk.
For six sigma limits, use the following formula instead as it was difficult to get a perfect match for the bin value. Also,
limit the number of bins you create as the more refined it is in decimal places the more number of six sigma matched results it will give. I did this by limiting to 50 bin arrays.
=IF(ROUND(E68,2)=ROUND($F$63,2),$F$65,IF(ROUND(E68,2)=ROUND($F$64,2),$F$65,0))
Could you explain your formula ?
Upper limit /lower limit should be Aritmetic mean +/- 3X standard deviation and not 6 times
I think, Both are correct. Its just telling deviation from the mean.
Thanks for the fabulous tutorial. I it's really handy .can u tell me how we can do process capability for defective variables ?
Can anyone explain what the "Bins" are?
Well explained. Thank you!
The calculation is a Ppk, not a Cpk. CpK is calulated based on a sub group average and estimated standard deviation of that subgroup.
Hi, obtaining the 6 sigma SL's become very difficult as the precision of measurement increases. Say when you are working in inches, and your measurements range in 3 decimal digits, say .235", the bins would have more than a single cell which can round off close to .235" value. This will result in multiple SL's line in the chart. Also, it is best to take min of LSL or LSL of 6sigma and max of USL or USL of 6 sigma to arrive at the bins value. If the mean offset from SL's, then you cant have the tolerance limits in the chart.
Let us know you have a solution for the above.
you are a genius, excellant
Thanks for this video!
Thank you, very helpful.
Very good presentation
Very nicely exlained
The CPK equation has a typo...the comma should be used but your showing the semicolon in between. Not sure how it worked on your demonstration?
check the below link.
www.excel-exercise.com/comma-or-semicolon-in-excel-formula/
I don't know what I am doing wrong but I am trying to calculate 3 sigma and it just is not working using this excel video, I am getting all zero's for 3 sigma
fantastic !
What are the bins? You shouldn’t calculated the distribution with the data? why with the bins
nice VDO ^^
How to get the "bin" value?
can you please share with us this excel file ?
Aaallll right
Thank you
How can I get a template? Please and Thanks :D
Good
Its ±3 sigma, else you get 12 sigma, not 6 sigma
good
Why Bins = 1;1,01;1,02... Please :(
loi nguyen van when you graph it youll understand
How do you calculate the UTL?
Hi, the UTL and LTL are set by specifications that you are measuring against :)
This tool does not work for my situation, I have product failure related projects, but there are no data. I also learned that the statistics works with data, if no pertinent data statistics would not work. If you are trying to use this technique in a manufacturing environment, this may not yield anything, for sure it is your money you have the right spend the way you wanted it. There are better ways to handle problems like this.
Mr. Uday Nair, Yes, you are very correct, attributes of failures are many, like some of them mentioned above and it would be very difficult to pick one of the attributes and collect data from the manufactured parts and detect the cause of failure, after doing all those exercise there is no defect found, and you may realize to look for some other characteristics and keep doing the exercise until you find a margin of error in one of the attributes. By the time you do the exercise to make the Six sigma applicable, few things are going to happen 1. You spend lot of time in collecting the data and analysis until a defect is detected, 2. Cost, cost of inspection process or Six Sigma process would exceed the cost of the product itself 3. The defect identified may not be a critical one or may be minor one and it would be acceptable as is and 4. What if the sample size or production lot is very few, not enough samples to draw data, what would be the next action? Not to mention that you would pass the delivery date of the product, your customers are waiting to get the parts to proceed with the next operation, the bottom line is the risk is much more than the benefits. I went through these exercises earlier and from the experience I am speaking all these.
Thank you for bringing your suggestions for discussion sake and hope this would shed some light in the subject matter we are in discussion and it may help some readers of this post. You or any one can contact me to know more about the process control, do not get me wrong this Six Sigma process is a good tool or technique in a product or process where there are no basic or datum or for some developmental projects to define the product or process. Contact: bmatthewbob@gmail.com
@Uday Nair Mr. Uday Nair, Yes, you are very correct, attributes of failures are many, like some of them mentioned above and it would be very difficult to pick one of the attributes and collect data from the manufactured parts and detect the cause of failure, after doing all those exercise there is no defect found, and you may realize to look for some other characteristics and keep doing the exercise until you find a margin of error in one of the attributes. By the time you do the exercise to make the Six sigma applicable, few things are going to happen 1. You spend lot of time in collecting the data and analysis until a defect is detected, 2. Cost, cost of inspection process or Six Sigma process would exceed the cost of the product itself 3. The defect identified may not be a critical one or may be minor one and it would be acceptable as is and 4. What if the sample size or production lot is very few, not enough samples to draw data, what would be the next action? Not to mention that you would pass the delivery date of the product, your customers are waiting to get the parts to proceed with the next operation, the bottom line is the risk is much more than the benefits. I went through these exercises earlier and from the experience I am speaking all these. This is from your note “of course not saying this will be conclusive but can be helpful”, how much helpful? Spending the amount of time and effort would justify the cost of your finding, a very minor one; one of my instructors told me that Statistics is value neutral, no product or process knowledge.
Thank you for bringing your suggestions for discussion sake and hope this would shed some light in the subject matter we are in discussion and it may help some readers of this post. You or any one can contact me to know more about the process control.
Do not get me wrong this Six Sigma process is a good tool or technique in a product or process where there are no basics or datum or for some developmental projects to define the product or process.
Contact: bmatthewbob@gmail.com.
Regards.
good