TOPIC: Six Sigma

I would like to ask you about a statement in a recent newsletter that I received on 5/2/2014. The article was describing Six Sigma. My question concerns a statement that said: "So much for the Six Sigma Methodology. Six sigma is also a statistical term, which identifies three standard deviations above and below the mean."

Is that accurate, or does six sigma identify six standard deviations above and below the mean? (Based on the text that followed the statement, I believe it is six standard deviations, but I would like to confirm that.)

Also, I would like to let you know how much I appreciate these newsletters. I find that every one either teaches me something, reinforces something that I knew, or simply helps boost my confidence when I realize I was able to identify the correct answer to a tough question. Thank you!
__________________________________
Lynn Walters

Lynn,

First of all I'm glad that you like our newsletter. We try to make it interesting and educational. And based on what you say we have achieve our goal...

And oh my! You are so right about what you noticed in that sentence from the newsletter... I did not write that clearly enough...

Here is a better explanation:

Six Sigma as a methodology:

Six Sigma is a disciplined, data-driven approach and methodology that was originally developed by Motorola. The idea was to improve processes by eliminating defects. And a defect is defined as "nonconformity of a product or service to its specifications." Six Sigma at many organizations simply means a measure of quality that strives for near perfection.

Six sigma as a measure of standard deviation:

The Six Sigma methodology has a strong focus on numbers, describes quantitatively how a process is performing and uses standard deviations (or sigma) from the mean as a measure. For instance you might read that "Your process is performing to six sigma". This means that 99.99% of your output is without defect.

For the PMP exam you have to know the statistical percentages for 1, 2, 3 and 6 sigma:

1 sigma: 68.27% (68.2689492…)
2 sigma: 95.45% (95.4499736…)
3 sigma: 99.73% (99.7300204…)
6 sigma: 99.99% (99.9999998027…)

So you are absolutely correct in saying "six sigma" means six standard deviations above and below the mean. I will change the text in that newsletter so we don't give the wrong impression.

Thank you, Cornelius. I really appreciate your taking the time to answer my question.

Would it be possible to elaborate a little more on the relation between six sigma quality requirement and process duration?

I have a question that results 13 days of the PERT task duration estimate and 1 day as a standard deviation. The correct answer reads that the task can be completed in 7 to 19 days period. I don't really understand where 6 days duration deviation came from...

Any input is greatly appreciated.

Would it be possible to elaborate a little more on the relation between six sigma quality requirement and process duration?

I have a question that results 13 days of the PERT task duration estimate and 1 day as a standard deviation. The correct answer reads that the task can be completed in 7 to 19 days period. I don't really understand where 6 days duration deviation came from...

Any input is greatly appreciated.

Hi Michael,

I think that you're keying in on the word "sigma" and that it is perhaps the cause of some confusion.

Generally when you hear "Six Simga" mentioned, it is as Cornelius says -- it is a quality managmenet methodology. Think of it as just one more method in a family of quality managment approaches over they years. Other examples would be TQM (Total Quality Management), or ZD (Zero Defects).

The idea behind Six Sigma is that they wanted to drive their defect rate down to near zero. The goal was to achieve a rate of approximagtely one defect per million items. If you look at a normal distribution (a bell curve), to achieve a probability at such a low level you need to go approximately six standard deviations away from the mean. That is where the name Six Sigma comes from. It's just a shorthand way of referencing their goal of defects being at a rate comparable with one in a million.


WIth respect to the PERT task duration, the beta distribution version of the formula basically is taking your date estimates and forcing them onto a simple probability function that looks like a bell curve.
The beta distribution PERT formula is: duration = (optimistic + pessimitic + 4*most likely) / 6
When using this simple model, the standard deviation (sigma) is equal to (optimistic - pessimistic) / 6.


You'll need to look at the context to know whether they're talking about the quality methodology or whether they are actually referencing six standard deviations away from the mean. However, if they intend the quality methodology then Six Sigma is generally capitalized.

With respect to your specific question, I'm afraid that I'm unable to speak to the details without seeing the details of the question. However, I hope that the above discussion helps you understand Six Sigma vs. PERT.

Steve, thank you for your input.
Just to clarify the question I am struggling with...

The activity has an optimistic estimate of 10 days, pessimistic estimate of 16 days and most likely estimate of 13 days. If your company has a quality requirement of Six Sigma what is the duration within which task must be completed?

I calculated 13 days mean.. and standard deviation = 1. Correct answer is 7 - 19 days. I don’t understand why it is +- 6 days from the mean.

Thanks.

Michael,

I agree with your calculation that using the PERT beta distribution you would come up with a 13 day mean and a standard deviation of 1 day. I appears, however, that the question intends you to take +/- 6 standard deviations (six sigma) from the mean. This produces the answer of 7 to 19 days.

However, I personally would have found this question a bit frustrating. If I see "a quality requirement of Six Sigma", then because it is capitalized then I typically would interpret it to mean that they are following the whole Six Sigma quality philosophy / framework -- and not necessarily that they intend you to take +/- 6 standard deviations in your calculations.

The bottom line is that I would view that question as poorly worded and a little unclear. Remember that the PMP exam is multiple choice. Since it can be interpreted two ways, I would do both methods and check to see if both answers appear as options. If only one answer is there, then you know which to pick. If both answers appear as options, then you'll just have to choose whichever interpretation makes most sense to you.

Dear Fresh,
I am sure the standard deviation was given or not described as pessimist or optimist viewpoint.. Probably they have not given as 3 point PERT estimation. That's why your standard deviation was 16-10=6

your question says 6 sigma is the target so it should be SDX6. = 6